Detect when write_file / patch calls fail during a turn and are never
superseded by a successful write to the same path. When the final
text response is delivered, append an advisory footer listing the
files that did NOT change — so models that over-claim 'patched 5 files'
after 4 silent failures can't hide the lie.
Catches the failure mode reported in Ben Eng's llm-wiki session:
grok-4.1-fast issued batches of parallel patches, half failed with
'Could not find old_string', and the agent summarised the turn
claiming every file was edited. The user had to manually run
'git status' each turn to catch it.
The verifier is a pure post-hoc check on tool results — no new LLM
calls, no synthetic messages injected into history (prompt cache
preserved), no changes to tool argument dispatch. Per-turn state is
keyed by path; a later successful write to the same path clears the
failure entry so single-file retry recovery is not flagged.
Wired into both _execute_tool_calls_concurrent and
_execute_tool_calls_sequential, so batched parallel patches and one-at-
a-time edits are both covered. Footer emission happens after the
agent loop exits, before transform_llm_output / post_llm_call plugin
hooks run, so plugins still see (and can modify) the augmented text.
Config: display.file_mutation_verifier (bool, default true) +
HERMES_FILE_MUTATION_VERIFIER env override.
31 unit tests in tests/run_agent/test_file_mutation_verifier.py cover
target extraction (write_file, patch-replace, patch-v4a single and
multi-file), error-preview extraction (JSON .error field and plain
string), per-turn state transitions (first-error-wins on repeated
failure, success supersedes failure), footer rendering (truncation
at 10 entries, user-actionable hint), and env/config precedence.
Companion docs updated: user-guide/configuration.md +
reference/environment-variables.md.
* feat(security): supply-chain advisory checker + lazy-install framework + tiered install fallback
Three coordinated mitigations for the Mini Shai-Hulud worm hitting
mistralai 2.4.6 on PyPI (2026-05-12) and for the next single-package
compromise that follows.
# What this PR makes true
1. Users with the poisoned mistralai 2.4.6 in their venv get a loud
detection banner with copy-pasteable remediation steps the moment
they run hermes (and on every gateway startup).
2. One quarantined / yanked PyPI package can no longer silently demote
a fresh install to 'core only' — the installer keeps every other
extra and tells the user which tier landed.
3. Future opt-in backends (Mistral, ElevenLabs, Honcho, etc.) can
lazy-install on first use under a strict allowlist, instead of
eagerly pulling everything at install time.
# Detection: hermes_cli/security_advisories.py
- ADVISORIES catalog (one entry currently: shai-hulud-2026-05 for
mistralai==2.4.6). Adding the next one is a single dataclass.
- detect_compromised() uses importlib.metadata.version() — no pip
dependency, works in uv venvs that lack pip.
- Banner cache (~/.hermes/cache/advisory_banner_seen) rate-limits
the startup banner to once per 24h per advisory.
- Acks persisted to security.acked_advisories in config.yaml; never
re-banner after ack.
- Wired into:
* hermes doctor — runs first, prints full remediation block
* hermes doctor --ack <id> — dismisses an advisory
* cli.py interactive run() and single-query branches — short
stderr banner pointing at hermes doctor
* gateway/run.py startup — operator-visible warning in gateway.log
# Lazy-install framework: tools/lazy_deps.py
- LAZY_DEPS allowlist maps namespaced feature keys (tts.elevenlabs,
memory.honcho, provider.bedrock, etc.) to pip specs.
- ensure(feature) installs missing deps in the active venv via the
uv → pip → ensurepip ladder (matches tools_config._pip_install).
- Strict spec safety regex rejects URLs, file paths, shell metas,
pip flag injection, control chars — only PyPI-by-name accepted.
- Gated on security.allow_lazy_installs (default true) plus the
HERMES_DISABLE_LAZY_INSTALLS env var for restricted/audited envs.
- Migrated three backends as proof of pattern:
* tools/tts_tool.py — _import_elevenlabs() calls ensure first
* plugins/memory/honcho/client.py — get_honcho_client lazy-installs
* tts.mistral / stt.mistral entries pre-registered for when PyPI
restores mistralai
# Installer fallback tiers
scripts/install.sh, scripts/install.ps1, setup-hermes.sh:
- Centralised _BROKEN_EXTRAS list (currently: mistral). Edit one
array when a transitive breaks; users keep every other extra.
- New 'all minus known-broken' tier between [all] and the existing
PyPI-only-extras tier. Only kicks in when [all] fails resolve.
- All three tiers explicit: every fallback announces which tier
landed and prints a re-run hint when not on Tier 1.
- install.ps1 and install.sh both regenerate their tier specs from
the same _BROKEN_EXTRAS array so updates stay in sync.
Side effect: install.ps1 Tier 2 spec previously hardcoded 'mistral'
in its extra list — bug fixed by the refactor (mistral is filtered
out).
# Config
hermes_cli/config.py — DEFAULT_CONFIG.security gains:
- acked_advisories: [] (advisory IDs the user has dismissed)
- allow_lazy_installs: True (security gate for ensure())
No config version bump needed — both keys nest under existing
security: block, and load_config's deep-merge picks up DEFAULT_CONFIG
defaults for users with older configs.
# Tests
tests/hermes_cli/test_security_advisories.py — 23 tests covering:
- detect_compromised matches/non-matches, wildcard frozenset
- ack persistence, idempotence, blank rejection, config-failure path
- banner cache rate limiting + 24h re-banner + ack-stops-banner
- short_banner_lines / full_remediation_text / render_doctor_section /
gateway_log_message
- shipped catalog well-formedness invariant
tests/tools/test_lazy_deps.py — 40 tests covering:
- spec safety: 11 safe parametrized + 18 unsafe parametrized
- allowlist: unknown-feature rejection, namespace.name shape,
every shipped spec passes the safety regex
- security gating: config flag, env var, default, fail-open
- ensure() happy/sad paths: already-satisfied, install success,
pip stderr surfaced on failure, install-succeeds-but-still-missing
- is_available, feature_install_command
Combined: 63 new tests, all passing under scripts/run_tests.sh.
# Validation
- scripts/run_tests.sh tests/hermes_cli/test_security_advisories.py
tests/tools/test_lazy_deps.py → 63/63 passing
- scripts/run_tests.sh tests/hermes_cli/test_doctor.py
tests/hermes_cli/test_doctor_command_install.py
tests/tools/test_tts_mistral.py tests/tools/test_transcription_tools.py
tests/tools/test_transcription_dotenv_fallback.py → 165/165 passing
- scripts/run_tests.sh tests/hermes_cli/ tests/tools/ →
9191 passed, 8 pre-existing failures (verified on origin/main
before this change)
- bash -n on install.sh and setup-hermes.sh → OK
- py_compile on all modified .py files → OK
- End-to-end smoke test of detect_compromised + render_doctor_section
+ gateway_log_message with mocked installed version → produces
copy-pasteable remediation output
# Community
Full advisory + remediation steps:
website/docs/community/security-advisories/shai-hulud-mistralai-2026-05.md
Short-form post drafts (Discord, GitHub pinned issue, README banner):
scripts/community-announcement-shai-hulud.md
Refs: PR #24205 (mistral disabled), Socket Security advisory
<https://socket.dev/blog/mini-shai-hulud-worm-pypi>
* build(deps): pin every direct dep to ==X.Y.Z (no ranges)
Companion to the supply-chain advisory work: replace every >=/</~= range
in pyproject.toml's [project.dependencies] and [project.optional-dependencies]
with an exact ==X.Y.Z pin sourced from uv.lock.
Why: ranges allow PyPI to ship a fresh version of any direct dep at any
time without a code review on our side. With ranges, the malicious
mistralai 2.4.6 release would have been pulled by every fresh
'pip install -e .[all]' for the hours between upload and PyPI's
quarantine — exactly the install window we got hit on. Exact pins close
that window: the only way a new package version reaches a user is via
an intentional update on our end.
What the user-facing change is: nothing, behavior-wise. Every package
resolves to the same version it was already resolving to via uv.lock —
the pins just remove the resolver's freedom to pick a different one.
Cost: any user installing Hermes alongside another package that requires
a newer pin gets a resolver conflict. Acceptable for our isolated-venv
install path; documented in the new comment block.
Build-system requires line (setuptools>=61.0) is intentionally left
as a range — pinning the build backend would block fresh pip from
bootstrapping the build on architectures where that exact wheel isn't
available.
mistral extra (mistralai==2.3.0) is pinned but stays out of [all]
(per PR #24205). 'uv lock' regeneration will fail until PyPI restores
mistralai; lockfile regeneration is gated behind that, NOT on every PR.
LAZY_DEPS in tools/lazy_deps.py also moved to exact pins so the lazy-
install pathway can never resolve a different version than the one
declared in pyproject.toml.
Validation:
- Cross-checked all 77 pinned direct deps in pyproject.toml against
uv.lock — every pin matches the resolved version exactly.
- Cross-checked all LAZY_DEPS specs against uv.lock — same.
- 'uv pip install -e .[all] --dry-run' resolves 205 packages cleanly.
- tests/tools/test_lazy_deps.py + tests/hermes_cli/test_security_advisories.py
→ 63/63 passing (every shipped spec passes the safety regex).
- Doctor + TTS + transcription targeted suite → 146/146 passing.
* build(deps): hash-verify transitives via uv.lock; remove unresolvable [mistral] extra
You asked: 'what about the dependencies the dependencies rely on?' —
correctly noting that exact-pinning direct deps in pyproject.toml does
NOT cover the transitive graph. `pip install` and `uv pip install` both
re-resolve transitives fresh from PyPI at install time, so a compromised
transitive (e.g. `httpcore` if it got worm-poisoned tomorrow) would
still hit our users even with every direct dep exact-pinned.
# What this commit fixes
1. **Both real installer scripts now prefer `uv sync --locked` as Tier 0.**
uv.lock records SHA256 hashes for every transitive — a compromised
package with a different hash gets REJECTED. Falls through to the
existing `uv pip install` cascade if the lockfile is missing or
stale, with a loud warning that the fallback path does NOT
hash-verify transitives. Previously only `setup-hermes.sh` (the dev
path) used the lockfile; `scripts/install.sh` and `scripts/install.ps1`
(the paths fresh users actually run) skipped it.
2. **Removed the `[mistral]` extra entirely.** The `mistralai` PyPI
project is fully quarantined right now — every version returns 404,
so any pin we wrote was unresolvable, which broke `uv lock --check`
in CI. Restoration is documented in pyproject.toml as a 5-step
checklist (verify, re-add extra, re-enable in 4 modules, regenerate
lock, optionally re-add to [all]).
3. **Regenerated uv.lock.** 262 packages, mistralai/eval-type-backport/
jsonpath-python pruned. `uv lock --check` now passes.
# Defense-in-depth view
| Layer | Where | Protects against |
|----------------------------|-------------------|-------------------------------------------|
| Exact pins in pyproject | direct deps | new mistralai 2.4.6-style direct compromise |
| uv.lock + `--locked` install | transitive graph | transitive worm injection |
| Tier-0 hash-verified path | install.sh / .ps1 | actually USE the lockfile in fresh installs |
| `uv lock --check` CI gate | every PR | drift between pyproject and lockfile |
| `hermes_cli/security_advisories.py` | runtime | cleanup for users who already got hit |
The exact pinning + hash verification together close the supply-chain
gap. Without the lockfile path, exact pins alone are theater.
# Validation
- `uv lock --check` → passes (262 packages resolved, no drift).
- `bash -n` on install.sh + setup-hermes.sh → OK.
- 209/209 tests passing across new + adjacent test files
(test_lazy_deps.py, test_security_advisories.py, test_doctor.py,
test_tts_mistral.py, test_transcription_tools.py).
- TOML parse OK.
* chore: remove community announcement drafts (PR body covers it)
* build(deps): lazy-install every opt-in backend (anthropic, search, terminal, platforms, dashboard)
Extends the lazy-install framework to cover everything that's not used by
every hermes session. Base install drops from ~60 packages to 45.
Moved out of core dependencies = []:
- anthropic (only when provider=anthropic native, not via aggregators)
- exa-py, firecrawl-py, parallel-web (search backends; only when picked)
- fal-client (image gen; only when picked)
- edge-tts (default TTS but still optional)
New extras in pyproject.toml: [anthropic] [exa] [firecrawl] [parallel-web]
[fal] [edge-tts]. All added to [all].
New LAZY_DEPS entries: provider.anthropic, search.{exa,firecrawl,parallel},
tts.edge, image.fal, memory.hindsight, platform.{telegram,discord,matrix},
terminal.{modal,daytona,vercel}, tool.dashboard.
Each import site now calls ensure() before importing the SDK. Where the
module had a top-level try/except (telegram, discord, fastapi), the
graceful-fallback pattern was extended to lazy-install on first
check_*_requirements() call and re-bind module globals.
Updated test_windows_native_support.py tzdata check from snapshot
(>=2023.3 literal) to invariant (any version + win32 marker).
Validation:
- Base install: 45 packages (was ~60); 6 newly-extracted packages absent
- uv lock --check: passes (262 packages, no drift)
- 209/209 lazy_deps + advisory + doctor + tts/transcription tests passing
- py_compile clean on all 12 modified modules
The `mistralai` PyPI package was quarantined on 2026-05-12 after a
malicious 2.4.6 release. Every fresh resolve (AUR makepkg, Docker build,
CI run, install.sh first-run) currently fails on
`mistralai>=2.3.0,<3` because PyPI returns zero candidates.
Existing users running `hermes update` mostly didn't notice — `hermes
update` falls back from `.[all]` to per-extra retries and silently
skips mistral with a warning that scrolls past. But fresh installs
hard-fail or lose every other extra.
Changes:
- pyproject.toml: drop `hermes-agent[mistral]` from `[all]` and
`[termux-all]`. The `mistral` extra itself is preserved so users
can opt back in once PyPI un-quarantines.
- hermes_cli/tools_config.py: hide Mistral Voxtral TTS from the
`hermes tools` provider picker until restored.
- hermes_cli/web_server.py: drop "mistral" from dashboard STT options.
- tools/transcription_tools.py: explicit `provider: mistral` returns
"none" with a clear status message; auto-detect skips mistral.
- tools/tts_tool.py: dispatcher returns a clear "temporarily disabled"
error before any SDK import attempt (avoids cached-stale-package
surprises).
- tests/tools/: update three test files to assert the new disabled
behavior. Each test docstring records why and points at the rollback
trigger (PyPI un-quarantines mistralai).
Restore plan: revert this commit once the package is available on PyPI
again. The behavior change is intentional and documented in code
comments + test docstrings to make the rollback trivial.
Validation:
- scripts/run_tests.sh tests/tools/ -k 'mistral or stt or tts' →
425/425 passing.
Refs: https://pypi.org/simple/mistralai/ (currently
"pypi:project-status: quarantined").
Handle MiniMax OAuth expiry values consistently across CLI and dashboard
flows, fix CLI status/add behavior, and force pooled OAuth runtime
requests through Anthropic Messages.
- web_server._minimax_poller: parse expired_in via the shared resolver
so unix-ms absolute timestamps stop landing as TTL seconds and crashing
with 'year 583911 is out of range' when a user connects MiniMax OAuth
from the dashboard.
- auth._minimax_oauth_login / _refresh_minimax_oauth_state: same fix on
the CLI login + refresh paths.
- auth.get_auth_status: dispatch minimax-oauth to its dedicated status
function instead of falling through.
- auth_commands.auth_add_command: 'hermes auth add minimax-oauth' now
starts the device-code login flow and persists a pool entry with the
access + refresh tokens, instead of requiring credentials to already
exist.
- runtime_provider._resolve_runtime_from_pool_entry: pin pooled
minimax-oauth credentials to anthropic_messages so a stale
model.api_mode: chat_completions can't send requests to
/anthropic/chat/completions and trigger MiniMax nginx 404s.
Co-authored-by: Cursor <cursoragent@cursor.com>
Qwen models on Nous Portal (e.g. qwen3.6-plus) now get the same envelope-layout
cache_control markers and long-lived (1h cross-session) cache treatment as
Portal Claude. Portal proxies to OpenRouter with identical wire-format and
cache_control semantics, but the prior policy left Portal Qwen falling through
to the alibaba-family branch (which only matches provider=opencode/alibaba),
serving 0% cache hits and re-billing the full prompt every turn.
Scope is narrow: Portal Claude OR Portal Qwen. Other models on Portal keep
their existing behavior.
- _anthropic_prompt_cache_policy: add (is_nous_portal and qwen) -> (True, False)
- _supports_long_lived_anthropic_cache: drop Claude-only gate for Portal so
Qwen also gets the validated 1h cross-session layout
- tests cover both functions, both bare and vendored qwen slug forms, and
the rejection of non-Claude non-Qwen Portal traffic
Free-tier users were seeing 'No free models currently available.' in the
`hermes model` and post-login pickers even though qwen/qwen3.6-plus is
free on the Portal right now. Three independent breakages compounded:
1. The docs-hosted catalog manifest at website/static/api/model-catalog.json
was not regenerated when _PROVIDER_MODELS['nous'] was updated, so users
fetching the manifest got a list that didn't include qwen/qwen3.6-plus.
2. _resolve_nous_pricing_credentials() returned ('', '') on any auth blip,
collapsing get_pricing_for_provider('nous') to {} and making every
curated model fall through the free-tier filter as 'paid'.
3. Even with healthy pricing, the picker only ever showed models from the
in-repo curated list intersected with live pricing — a Portal-flagged
free model not yet in the curated list could never appear.
Changes:
- hermes_cli/models.py: new union_with_portal_free_recommendations() that
augments the curated list with Portal freeRecommendedModels entries
(with synthetic free pricing so partition keeps them). The Portal's
/api/nous/recommended-models endpoint is now the source of truth for
free-tier surfacing — old Hermes builds will see new free models
without a CLI release.
- hermes_cli/models.py: _resolve_nous_pricing_credentials() falls back to
the public inference base URL when runtime cred resolution fails.
The /v1/models endpoint exposes pricing without auth, so silently
returning {} just because a refresh token expired was wrong.
- hermes_cli/auth.py + hermes_cli/main.py: both free-tier picker call
sites call union_with_portal_free_recommendations() before partition.
- tests/hermes_cli/test_models.py: 7 tests covering union behaviour
(prepend, dedup, end-to-end with stale pricing, empty/missing/error
payloads, invalid entries).
- tests/hermes_cli/test_model_catalog.py: drift guard
TestManifestMatchesInRepoLists fails CI when _PROVIDER_MODELS['nous']
or OPENROUTER_MODELS is edited without re-running
scripts/build_model_catalog.py. Verified empirically that removing a
manifest entry triggers an assertion with an actionable error message.
Validation:
- 133/133 targeted tests pass (test_models, test_model_catalog,
test_auth_nous_provider).
- Live E2E against the real Portal:
- Stale curated list ['claude-opus','claude-sonnet','gpt-5.4'] (no
qwen) → after union: ['qwen/qwen3.6-plus', ...] →
partition(free_tier=True): selectable=['qwen/qwen3.6-plus'].
- Simulated expired refresh token → anon fetch returns 403 pricing
entries including qwen/qwen3.6-plus -> {prompt:0, completion:0}.
- ruff: clean.
cua-driver was only installed once on toolset enable: `_run_post_setup` early-returns when the binary is already on PATH, so upstream fixes (e.g. v0.1.6 Safari window-focus fix) never reached existing users without manual reinstall.
Two refresh points now:
- `hermes update` re-runs the upstream installer at the end of the update if cua-driver is on PATH (macOS-only, no-op otherwise). Ties driver freshness to the user-controlled update cadence — no startup latency, no per-launch GitHub API call.
- `hermes computer-use install --upgrade` for manual force-refresh.
The upstream `install.sh` always pulls the latest release, so re-running is the canonical upgrade path. No version-comparison logic needed.
`hermes computer-use status` now shows the installed version, and points at `--upgrade` for refreshing.
Fixes#22832.
## Root cause
`hermes_cli/web_server.py:start_oauth_login` dispatched OAuth flows by
the catalog's `flow` field rather than provider id:
if catalog_entry["flow"] == "pkce":
return _start_anthropic_pkce()
The catalog had two `flow: "pkce"` entries — `anthropic` and
`minimax-oauth` — so clicking "Login" on MiniMax in the dashboard's
Keys tab unconditionally launched the Anthropic/Claude PKCE flow.
## Fix
Three changes in `hermes_cli/web_server.py`:
1. Catalog entry for `minimax-oauth` changed from `flow: "pkce"` to
`flow: "device_code"`. From a UX perspective MiniMax is a
verification-URI + user-code flow (open URL, enter code, backend
polls) — same shape as Nous's device-code flow. The PKCE bit
(verifier + challenge from `_minimax_pkce_pair`) is a security
extension that doesn't change the operator experience; the existing
dashboard modal already renders `device_code` correctly for this UX.
2. New MiniMax branch in `_start_device_code_flow`, mirroring the
existing Nous branch but calling MiniMax-specific helpers
(`_minimax_request_user_code`, `_minimax_pkce_pair`). Stashes
verifier + state in the session for the poller to consume. Handles
the overloaded `expired_in` field (could be unix-ms timestamp OR
seconds-from-now duration) the same way `_minimax_poll_token` does.
3. New `_minimax_poller` background thread mirroring `_nous_poller`.
Calls `_minimax_poll_token` → on success builds the same
`auth_state` dict the CLI flow (`_minimax_oauth_login`) builds, and
persists via `_minimax_save_auth_state` so the dashboard path leaves
the system in the same state as `hermes auth add minimax-oauth`.
Plus a dispatcher tightening to prevent regression: the `pkce` branch
now requires `provider_id == "anthropic"`, so any future PKCE provider
added without a proper start function gets a clean
`400 Unsupported flow` rather than silently launching Anthropic OAuth.
## Test
New `tests/hermes_cli/test_web_oauth_dispatch.py`:
- Regression test asserting MiniMax start does NOT return claude.ai
- Sanity test that Anthropic PKCE still works after the dispatcher
tightening
- Forward-looking test: a hypothetical pkce-flagged provider without
an explicit branch is rejected cleanly rather than misrouted
## Limitations
- The dashboard MiniMax path defaults to `region="global"`. CN-region
operators can still use the CLI flow which supports `--region cn`.
Adding a region toggle to the dashboard UI is a follow-up.
Follow-up to #23863 (CJK table alignment). The realigner was
correctly padding pipes to identical column offsets, but when a
table's natural width exceeds terminal cells it produced lines that
the terminal soft-wrapped mid-cell, destroying column alignment
visually even though the bytes were perfectly padded. Reported as
'columns are not aligned' on tables containing one long row alongside
several short rows.
Approach mirrors Claude Code's MarkdownTable.tsx narrow-terminal
fallback: when realign_markdown_tables is given an available_width
budget and the rebuilt horizontal table exceeds it, render each body
row as 'Header: value' lines separated by a thin ─ rule. Word-wraps
oversize values at the budget with a 2-space continuation indent.
- agent/markdown_tables.py: realign_markdown_tables(text, available_width=None);
threshold check at the top of _render_block flips into a new
_render_vertical fallback. Includes _wrap_to_width with hard-break
for tokens longer than the budget.
- cli.py: helper _terminal_width_for_streaming() returns
shutil.get_terminal_size().columns minus _STREAM_PAD and a 2-cell
safety margin; passed to all three realign call sites
(_render_final_assistant_content for strip+render Panel paths, and
the streaming flushers in _emit_stream_text / _flush_stream).
- tests/agent/test_markdown_tables.py: 4 new tests covering the
overflow-vertical fallback for ASCII + CJK content, the
'fits → keep horizontal' case, and the long-cell wrap with indent.
Live-verified: with COLUMNS=100, the user's reported 'long row in
ASCII table' case now renders as vertical key-value rows that all fit
the panel; the 6-column CJK comparison table still renders as an
aligned horizontal table because it fits inside 100 cols.
The old mtime-tracking staleness machinery (_tui_build_needed,
_hermes_ink_bundle_stale, _find_bundled_tui) tried to avoid rebuilding
by comparing source timestamps to dist/entry.js. This was fragile and
added ~100 lines of code. Replace with three clear paths:
1. HERMES_TUI_DIR set (prebuilt/nix): just node dist/entry.js, no build
2. --dev mode: tsx src/entry.tsx, no build, hot reload
3. Normal: always npm run build (esbuild is ~1s, correctness > caching)
Also error when HERMES_TUI_DIR is set with --dev (footgun: prebuilt
bundle has no source code to hot-reload).
Set HERMES_SESSION_ID using the existing session_context.py ContextVar
system for concurrency safety (multiple gateway sessions in one process
won't cross-talk). Also writes os.environ as fallback for CLI mode.
Touchpoints:
- gateway/session_context.py: Add _SESSION_ID ContextVar + _VAR_MAP entry
- run_agent.py: Set both ContextVar and os.environ at init and on
context-compression rotation
- tools/environments/local.py: Bridge ContextVars into subprocess env
in _make_run_env() (ContextVars don't propagate to child processes)
- tests/run_agent/test_session_id_env.py: 3 tests covering env, provided
ID, and ContextVar paths
execute_code subprocess already passes HERMES_* prefixed vars through
_scrub_child_env (line 82: _SAFE_ENV_PREFIXES includes 'HERMES_').
Primary use case: webhook-triggered agents that need to include a
`--resume <session_id>` takeover command in their output.
Cuts input cost for first-turn Claude requests by ~85-90% on subsequent
sessions within an hour. Tools array (~13k tokens for default toolset) +
stable system prefix (~5-8k tokens) get a 1h cache_control marker; the
volatile suffix (memory, USER profile, timestamp, session id) sits in a
separate non-cached block at the end so it doesn't poison the cross-session
prefix when it changes.
Provider gate: Claude on native Anthropic (incl. OAuth subscription),
OpenRouter, and Nous Portal (which proxies to OpenRouter). All other
providers keep today's system_and_3 layout unchanged.
Layout (4 cache_control breakpoints, Anthropic max):
1. tools[-1] -> 1h (cross-session)
2. system content[0] -> 1h (cross-session, stable prefix)
3. messages[-2] -> 5m (within-session rolling)
4. messages[-1] -> 5m (within-session rolling)
Within-session rolling shrinks from 3 messages to 2 to free the breakpoint
budget. On Claude with realistic tool loadouts the long-lived tier carries
the bulk of cross-session value anyway.
System prompt is now always assembled cache-friendly: stable identity /
guidance / skills / platform hints first, then session-stable context
files (AGENTS.md, .cursorrules), then per-call volatile content. Old
single-string callers see the same logical content (same join order),
just reordered so volatile lives at the end.
Config knobs (defaults shown):
prompt_caching:
cache_ttl: "5m" # rolling-window TTL (unchanged)
long_lived_prefix: true # opt-out switch
long_lived_ttl: "1h" # cross-session prefix TTL
Live E2E (tests/agent/test_prompt_caching_live.py, gated on
OPENROUTER_API_KEY) on anthropic/claude-haiku-4.5 with default toolset:
Call 1 (cold): cache_write=13,415 cache_read=0
Call 2 (NEW agent + msg): cache_write=391 cache_read=13,025
Cross-session reuse: 97.09%
Implementation:
* agent/prompt_caching.py: new apply_anthropic_cache_control_long_lived()
+ mark_tools_for_long_lived_cache(); existing apply_anthropic_cache_control()
preserved verbatim for the fallback path.
* agent/anthropic_adapter.py: convert_tools_to_anthropic() now forwards
cache_control onto each Anthropic-format tool dict.
* run_agent.py: _build_system_prompt_parts() returns the 3-tier dict;
_build_system_prompt() joins them (backward compatible).
_supports_long_lived_anthropic_cache() policy added next to the existing
_anthropic_prompt_cache_policy() (which now also recognises Nous Portal
Claude — pre-existing gap fixed in passing).
_build_api_kwargs() resolves tools_for_api once and propagates the
marker through all four build paths (anthropic_messages, bedrock,
codex_responses, profile/legacy chat completions).
Long-lived flag plumbed into the runtime snapshot/restore + model-switch
+ fallback-promotion paths.
Tests:
* tests/agent/test_prompt_caching.py: +8 tests (TestMarkToolsForLongLivedCache,
TestApplyAnthropicCacheControlLongLived).
* tests/run_agent/test_anthropic_prompt_cache_policy.py: +9 tests
(TestSupportsLongLivedAnthropicCache matrix across 8 endpoint classes
+ a fallback-target case).
* tests/agent/test_prompt_caching_live.py: new live E2E (skipif when
OPENROUTER_API_KEY is unset; runs outside the hermetic suite).
* Targeted suites: 327/327 pass (caching/adapter/policy/builder).
* tests/agent/ + tests/run_agent/: 3992 pass, 17 skip, 1 pre-existing
flake (test_async_httpx_del_neuter::test_same_key_replaces_stale_loop_entry,
verified failing on pristine origin/main).
#23482 fixed cache poisoning in the sync path: when a Codex auxiliary
timeout closes the underlying OpenAI client, _evict_cached_client_instance
walks CodexAuxiliaryClient wrappers via their _real_client attribute and
drops the cache entry so the next aux call rebuilds.
The cache key includes async_mode (see _client_cache_key), so the sync and
async clients for the same provider live in two distinct entries pointing
at the same underlying transport. The fix walked the sync wrapper's
_real_client correctly but the async wrappers
(AsyncCodexAuxiliaryClient, AsyncAnthropicAuxiliaryClient,
AsyncGeminiNativeClient) never exposed _real_client at all, so the async
entry survived eviction and kept handing out the poisoned client.
Effect on async aux callers: one timeout now poisons every subsequent
async aux call (compression, vision, session_search, title_generation)
with 'Connection error' until gateway restart -- even while the sync
route recovered as designed in #23482.
Mirror the sync wrapper's _real_client onto each async wrapper so the
existing eviction helper finds them. Three changes, one per wrapper:
- AsyncCodexAuxiliaryClient: self._real_client = sync_wrapper._real_client
(the underlying OpenAI client)
- AsyncAnthropicAuxiliaryClient: same shape
- AsyncGeminiNativeClient: self._real_client = sync_client (Gemini's
native facade is itself the leaf; no OpenAI client beneath it)
Update _evict_cached_client_instance docstring to reflect that it now
covers both sync and async wrappers via the same attribute walk.
Test: TestAuxiliaryClientPoisonedCacheEviction.test_evict_cached_client_instance_walks_async_wrapper
seeds both sync and async cache entries pointing at the same leaf and
asserts both are dropped on a single eviction call. Verified the test
fails without the wrapper changes ("async cache entry survived
eviction -- wrapper is missing _real_client") and passes with them.
Refs #23482, #23432
CJK and emoji glyphs render as two terminal cells but JS String#length
and the model's own padding count them as one, so any markdown table
with Chinese / Japanese / Korean cells drifts right per row when a
real terminal renders it. Both surfaces fix this with a display-cell
width measurement (wcswidth on the Python side, stringWidth on the
TUI side).
Changes:
- agent/markdown_tables.py: new helper. realign_markdown_tables(text)
detects markdown table blocks (header + |---| divider) and
rewrites the row padding using wcwidth.wcswidth so every pipe and
dash lines up across rows. No-op on text without tables.
- cli.py: hook the helper into _render_final_assistant_content for
strip / render modes (raw passes through untouched), and into the
streaming line emitter so live token-by-token rendering also
produces aligned tables. A small two-buffer state machine in
_emit_stream_text holds table rows until the block ends, then
flushes them through the realigner so all rows pad to a single
per-column width.
- ui-tui/src/components/markdown.tsx: renderTable now uses
stringWidth (Bun.stringWidth fast path + East-Asian-width-aware
fallback, already memoised in @hermes/ink) instead of UTF-16
String#length for both column-width measurement and per-cell
padding. Drops the comment that documented the bug as a deliberate
limitation.
Validation:
- New tests/agent/test_markdown_tables.py (11): every rebuilt block
shares pipe column offsets across rows for pure CJK, mixed
CJK+emoji, ragged-row, and multi-table inputs.
- Updated tests/cli/test_cli_markdown_rendering.py: the existing
strip-mode test asserted exact whitespace; rewritten to assert the
alignment contract (cell content survives + every rendered row
shares pipe offsets).
- New ui-tui markdown.test.ts case (1): rendered column-2 start
offset is identical for the header + every body row, including
the CJK row that drifted before the fix.
- Live: hermes chat -q with the user-reported screenshot prompt now
produces a perfectly aligned table on the wire (header, divider,
4 body rows including '通义千问', all pipes at identical columns).
The /model picker for Nous Portal users was returning the in-repo
_PROVIDER_MODELS["nous"] snapshot — which only updates on Hermes
releases — instead of the remote manifest published at
https://hermes-agent.nousresearch.com/docs/api/model-catalog.json.
OpenRouter already pulled from the manifest via fetch_openrouter_models;
"nous" was the only curated provider where the existing manifest
plumbing (get_curated_nous_model_ids → get_curated_nous_models) was
defined but not wired into the picker pipeline. Switch the curated
build in list_authenticated_providers to use it, with the same
graceful fallback to the in-repo snapshot when the manifest is
unreachable.
Test: tests/hermes_cli/test_model_catalog.py exercises the picker with
a patched manifest and asserts the manifest's nous list reaches
list_picker_providers. Falls-back-to-static path was already covered
by test_curated_nous_ids_falls_back_to_hardcoded_on_empty_catalog.
Follow-up to PR #23824. Adds two correctness fixes on top of the
contributor's salvaged commit:
1. Stale-dist fallback no longer gated on `fatal=False`. `cmd_dashboard`
passes `fatal=True` and is the primary scenario this fallback is for
(issue #23817 — Windows Scheduled Task at logon). The previous gate
meant the fallback never fired in the case it was designed for.
2. `--skip-build` now verifies the dist actually exists before starting
the server. Without this, a misconfigured pre-build would launch the
dashboard pointing at a missing dist and silently serve 404s. We now
exit 1 with a clear "pre-build first: cd web && npm run build"
message, and on success print which dist directory is being used.
Verified end-to-end on Linux:
- build fails + stale dist (fatal=True) -> fallback fires
- build fails + no dist (fatal=True) -> exit 1 with stderr surfaced
- build fails + stale dist (fatal=False) -> fallback fires
- --skip-build + missing dist -> exit 1 with clear guidance
- --skip-build + valid dist -> 'Skipping web UI build...'
On Windows systems using a Chinese GBK locale, `hermes update` could misreport the Web UI build as failed even when `npm run build` actually succeeded. The failure was caused by Python decoding captured npm output with the process locale inside a background subprocess reader thread. When npm emitted bytes such as `0x85`, decoding under GBK raised `UnicodeDecodeError`, and Hermes then surfaced a misleading "Web UI build failed" warning.
This change makes the npm install/npm ci path and the Web UI build step decode captured output explicitly as UTF-8 with `errors="replace"`. That keeps unexpected bytes from crashing output collection, preserves successful builds, and prevents false negatives during update on Windows.
The patch also adds regression tests that verify these subprocess calls always use explicit UTF-8 decoding with replacement semantics.
When the user's main provider is openai-codex on the ChatGPT-account
backend (https://chatgpt.com/backend-api/codex), sending a native image
attachment encodes it as data:image/...base64,... in the input_image
field. The OpenAI Responses API on the public endpoint accepts that, but
the ChatGPT-account variant rejects it with HTTP 400:
Invalid 'input[N].content[K].image_url'. Expected a valid URL, but got
a value with an invalid format.
Hermes' image-rejection phrase list didn't include this wording, so the
error escaped the strip-and-retry branch and fell through to the generic
recovery path: model fallback → context-too-large → compression cascade
→ auxiliary OpenRouter 402 spam (issue #23570).
Add a NARROW phrase keyed on the field-path apostrophe used by the Codex
Responses error format: "image_url'. expected". This matches the actual
error format without false-tripping on generic 'Expected a valid URL'
errors from unrelated tools (webhooks, redirect_uri, etc.). Once matched,
the existing branch strips images from history, sets _vision_supported=
False for the session, and retries text-only.
Refs #23570 (1 of 3 image-replay improvements; persistence rewrite to
store image PATHS instead of inlined base64 is a separate follow-up)
* Revert "fix(goals): force judge to use tool calls instead of JSON-text replies (#23547)"
This reverts commit a63a2b7c78.
* Revert "fix(goals): forward standing /goal state on auto-compression session rotation (#23530)"
This reverts commit 4a080b1d5a.
* Revert "feat(goals): /goal checklist + /subgoal user controls (#23456)"
This reverts commit 404640a2b7.
Adds the only #17873 category not covered by the in-flight PRs #17962
(briandevans, reverse shell + download-execute) and #7993 (SHL0MS,
credential reads + curl/wget exfiltration): sudo invocations that an
LLM-driven agent can drive without TTY interaction.
The agent has no TTY, so the sudo forms that succeed without human
involvement are those reading the password from stdin (`-S` / `--stdin`)
or via an askpass helper (`-A` / `--askpass`). The shell-launch (`-s`)
and list-privileges (`-a`) flags are also gated since they are
privilege-relevant invocations the agent can chain after acquiring the
password (e.g. read SUDO_PASSWORD from .env -> sudo -S -s -> root shell).
Plain `sudo cmd` (no flag) is TTY-bound and excluded.
Two patterns:
1. Direct flag: `\bsudo\b[^;|&\n]*?\s+(?:-s\b|--stdin\b|-a\b|--askpass\b)`
The lazy `[^;|&\n]*?` consumes flag-arguments without spanning
command separators, so `sudo -u root -S whoami` matches (a textbook
offensive form that a strict `(?:\s+-[^\s]+)*` "leading flags only"
pattern would have missed because `root` is a flag-value not a flag).
2. Combined short flags: `\bsudo\b[^;|&\n]*?\s+-[a-z]*[sa][a-z]*\b`
Catches packed forms like `sudo -nS id` where multiple flags share
a single `-X` token.
`_normalize_command_for_detection` lowercases input before pattern
matching (tools/approval.py:340), so case variants of S/s and A/a
collapse — both letter-pairs are gated since each is a privilege-
relevant invocation.
Tests: 21 new cases in TestDetectSudoStdin (12 positive covering all
flag-order permutations including herestring source and printf-piped
forms; 9 negative including TTY-bound `sudo whoami`, interactive
`sudo -i`, env-var reference `$SUDO_USER`, doc lookup `man sudo`,
package install, and the `pseudosudo` word-boundary edge case).
Empirical coverage: 11/11 attacks matched, 0/10 false positives.
Refs: #17873 category 4. Adjacent: #17962 (reverse shell + download-
execute), #7993 (credential reads + curl/wget exfiltration).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Fixes#9590: Block explicit sudo -S (stdin password mode) commands
when the SUDO_PASSWORD environment variable is not configured.
The attack vector: the LLM constructs 'echo guessedpass | sudo -S cmd'
to brute-force sudo passwords, iterates based on sudo's error output
('Sorry, try again'). The existing _transform_sudo_command only
injects -S when SUDO_PASSWORD exists; without it, the LLM's explicit
sudo -S must be treated as a guessing attempt.
Changes:
- Add _check_sudo_stdin_guard() in approval.py: detects sudo -S when
SUDO_PASSWORD is absent, anchored to command-start positions
(^ ; && || | etc.) to avoid false positives on literal text
- Integrate into check_all_command_guards() above yolo/mode=off so
the block is unconditional (like the hardline floor)
- Add 6 tests covering: detection, allow-list, SUDO_PASSWORD bypass,
integration with check_all_command_guards, yolo non-bypass,
container backend bypass
Adds a TestCheckSendMessage class with 7 focused tests pinning the
four passing conditions and the failure modes:
- HERMES_KANBAN_TASK grants access (the new branch)
- HERMES_KANBAN_TASK short-circuits before consulting
session_context or gateway.status (so workers don't depend on
those import paths being healthy)
- HERMES_SESSION_PLATFORM=telegram grants access
- HERMES_SESSION_PLATFORM=local falls through to gateway check
- is_gateway_running()=True grants access
- All signals absent → False
- gateway.status ImportError is swallowed → False
Pinning the short-circuit (test #2) is the load-bearing one — it
documents the contract that worker-side availability cannot regress
to depending on gateway-side state lookups.
When a kanban worker subprocess hits the iteration budget, the agent
loop strips tools and asks the model for a summary. The model cannot
call kanban_block itself at that point, so the process exits rc=0
without calling kanban_complete or kanban_block — a protocol violation
that the dispatcher detects as a fatal error, giving up after 1 failure
and stranding downstream tasks.
Fix: after _handle_max_iterations() returns, check HERMES_KANBAN_TASK
and call kanban_block with a reason describing the exhaustion. The
dispatcher then sees a clean block transition instead of a protocol
violation, and the task can be retried or escalated by a human.
Fixes [Bug] kanban-worker exits cleanly (rc=0) on iteration-budget
exhaustion without calling kanban_complete or kanban_block #23216
When an auxiliary provider returns HTTP 402 (credit / payment), every
subsequent compression / title-gen / session-search / vision call still
re-tried it as the FIRST entry in the chain — burning ~1 RTT to hit 402
again, then falling back. On a long Discord/LCM session that meant dozens
of doomed 402s per minute (issue #23570).
Add a per-process unhealthy-provider cache with a 10 min TTL. When any
caller observes a payment error against a provider, the label is marked
unhealthy and skipped by:
* _resolve_auto Step-1 (main provider use-as-aux path)
* _resolve_auto Step-2 (aggregator/fallback chain)
* _try_payment_fallback (used by call_llm/acall_llm on first 402)
Skip-logs are throttled to once per minute per label so a bursty session
doesn't spam agent.log. Entries auto-expire so a topped-up account
recovers without manual intervention. The cache is in-process only by
design — multi-profile users with different keys per profile must each
hit the 402 once.
Refs #23570
When the Discord typing API call fails (rate limit, network error, 403),
_typing_loop returns early but the stale task remains in _typing_tasks.
Subsequent send_typing calls see the stale entry and skip, leaving no
typing indicator for the rest of the agent invocation.
Add finally block to _typing_loop to always remove the task from
_typing_tasks on exit, whether from cancellation, error, or normal
completion. This allows send_typing to create a fresh task.
3 new tests in test_discord_send.py:
- Task removed after API error
- Typing restartable after failure
- stop_typing cleans up
A YAML parse error in ~/.hermes/config.yaml caused load_config() to print
one line to stdout (Warning: Failed to load config: ...) and silently fall
back to DEFAULT_CONFIG, dropping every user override (auxiliary providers,
fallback chain, model settings). Users only noticed when downstream
behavior misbehaved — see issue #23570 where a tab-indent error in the
auxiliary section caused aux fallback to use OpenRouter (depleted) instead
of the configured Codex/MiniMax chain.
Now: log at WARNING (so 'hermes logs' surfaces it), write a prominent line
to stderr, dedup on (path, mtime_ns, size) so concurrent loads don't spam,
and re-warn after the user edits the file. Both call sites (raw read +
merged load) route through the same helper.
Refs #23570
Salvages the three substantive low-severity fixes from Gutslabs' #1974
"misc bug fixes" bundle. The other 8 claims in that PR were either
already fixed on main with superior implementations (state lock,
firecrawl lazy import, fcntl/msvcrt guard, path normalization, schema
migrations) or did not survive review.
- run_agent: `_materialize_data_url_for_vision` uses
`NamedTemporaryFile(delete=False)`; if `base64.b64decode` raises on a
corrupt data URL the temp file would persist forever. Wrap the
write in try/except and `os.unlink` the temp on failure.
- gateway/session: `append_to_transcript` JSONL write had no error
handling, so disk-full / read-only-fs / permission errors crashed the
message handler. The SQLite write above is the primary store, so
swallow OSError on the JSONL fallback with a debug log.
- gateway/status: `_read_pid_record` reads `pid_path.read_text()` after
an `exists()` check; if the PID file is deleted between the two
calls (concurrent gateway restart) we hit an unhandled OSError.
Catch it and return None.
Adds a regression test for the tempfile cleanup; the other two paths
are defensive try/excepts on infrequent OSError that don't warrant
dedicated tests.
Co-authored-by: Teknium <127238744+teknium1@users.noreply.github.com>
- Implement tests for normalizing perpetual markets and DEXs.
- Validate JSON output for main commands including markets, candles, and review.
- Ensure environment variable resolution and dotenv file reading are covered.
- Test export functionality for market data with expected output structure.
agent.redact._REDACT_ENABLED is snapshotted at import time from
HERMES_REDACT_SECRETS env. Under xdist a prior test in the same worker
can flip it, so test_exec_command_output_is_redacted was order-dependent.
Pin it via monkeypatch like test_terminal_output_transform_still_runs_strip_and_redact does.
1. Quick command exec ran in the gateway process's full environment
without env sanitization or output redaction. A quick command like
"env" or "printenv" would leak all API keys, OAuth tokens, and
bot credentials to the messaging user.
Fix: apply _sanitize_subprocess_env() before exec and
redact_sensitive_text() on output before returning.
2. GatewayRunner._pending_messages was written on every interrupt
(lines 1331-1334) but never read or consumed anywhere. The actual
interrupt delivery uses adapter._pending_messages (a separate dict).
Removed the write-only accumulation to prevent unbounded growth.
Two new tests:
- tests/gateway/test_telegram_format.py
test_message_too_long_splits_into_continuations_not_silent_truncation:
asserts edit_message returns success=True with continuation_message_ids
populated and message_id pointing at the last continuation when
content exceeds MAX_MESSAGE_LENGTH (#19537). Replaces the original
fail-on-overflow assertion with the split-and-deliver contract.
- tests/gateway/test_stream_consumer.py
TestEditOverflowSplitAndDeliver.test_consumer_advances_message_id_on_split_and_deliver:
asserts the consumer side updates _message_id to the latest
continuation, clears _last_sent_text, and fires on_new_message when
the adapter reports a split-and-deliver result.
Surface ready tasks that nobody claims within a threshold (default
30 min) regardless of why. One identity-agnostic signal that catches:
- Operator typo'd the assignee
- Profile was deleted, leaving its tasks stranded
- External worker pool (Codex CLI lane, custom daemon) is down
- Dispatcher misconfigured (wrong board / wrong HERMES_HOME)
Today the dispatcher correctly skips these (no respawn loop, good)
but nothing surfaces the fact that operator-actionable work is
accumulating. The new `stranded_in_ready` rule does that without
requiring a manual lane registry — it reads the most recent ready-
transition event (`created` / `promoted` / `reclaimed` / `unblocked`)
and fires when (now - last_ready_ts) > threshold.
Severity escalates with age: warning at threshold, error at 2x,
critical at 6x. The cli_hint and reassign actions point operators
at the right next step.
Out of scope deliberately:
- Lane registry (#20157 closed) — this signal supersedes it.
- Pushing the diagnostic into messaging gateways — diagnostics
are pull-only via 'hermes kanban diagnostics' for now; gateway
push is a separate UX decision.
Tests: 10 new + 461 existing kanban tests pass. E2E verified end-
to-end via 'hermes kanban diagnostics --json' against a 2h-old
stranded task — surfaces as error severity with correct actions.
Live-tested on gemini-3-flash-preview the judge kept returning empty
or non-JSON content, tripping the consecutive-parse-failures auto-
pause. Free-form JSON output is hopeful; tool-call schemas are
enforced server-side by virtually every modern provider.
Two new tools the judge calls:
- submit_checklist(items) — Phase A, decompose
- update_checklist(updates, new_items, reason) — Phase B, evaluate
Both phases now call the auxiliary client with tool_choice forcing
the right tool. read_file remains for Phase B history inspection,
with the loop exiting only when update_checklist is called or the
read budget is exhausted (at which point read_file is dropped from
the toolbox and update_checklist is forced).
Robustness:
- _call_judge_with_tool_choice falls back tool_choice forced→required→
auto if the provider rejects a particular shape.
- If a fully-broken provider still returns content instead of a tool
call, the legacy JSON-text parsers stay around as a last-ditch
backstop so we never silently lose a checklist.
- _normalize_update_args replaces the JSON parser for the apply
layer; same 1-based→0-based conversion + terminal-status filter.
Live verification: same fizzbuzz goal that was hitting 'judge model
returned unparseable output 3 turns in a row' before now terminates
in 2 turns, all 11 items marked completed with item-specific
evidence, no auto-pause. Agent log shows
'produced 11 checklist items via tool call' instead of the JSON-
parse path.
Tests: 7 new cases for the tool-call path (Phase A success, Phase B
update only, Phase B read_file→update, JSON-content backstop,
empty-text item dropping, non-terminal status filter).
When run_agent's _compress_context fires mid-turn it ends the parent
session in SessionDB and creates a new continuation session with a
fresh session_id. The /goal state is keyed on session_id in
state_meta ("goal:<sid>"), so without forwarding the goal silently
disappears: _get_goal_manager() rebinds for the new session_id,
load_goal() returns None, mgr.is_active() is False, and the
continuation loop dies with no user-visible signal.
Fix: in the same SessionDB transaction block that creates the
continuation session, copy state_meta[goal:<old>] →
state_meta[goal:<new>] when present. No-op when the user has no
active goal. Logged at INFO so a stuck loop is debuggable.
Tests cover the round-trip via SessionDB and the no-op path.
Affects all three run-conversation surfaces (CLI, gateway, TUI
gateway) because _compress_context is the single rotation site.
Added tests/gateway/test_stream_consumer_draft.py with 11 tests
covering:
- Transport selection: auto+dm-supported -> draft; auto+group -> edit;
explicit edit; explicit draft on unsupported adapter -> edit;
MagicMock adapter -> edit (back-compat for the existing test suite).
- Happy path: DM stream animates draft frames with a single shared
draft_id, then finalizes via a regular adapter.send.
- Group fallback: drafts entirely skipped in non-DM chats.
- Failure fallback: send_draft returning success=False disables drafts
for the rest of the response.
- Draft_id lifecycle: consecutive responses use distinct ids; tool
boundaries bump the id so post-tool text animates fresh below the
tool-progress bubble (the openclaw #32535 leak guard).
- _already_sent contract: drafts must NOT set the flag so the gateway's
fallback final-send still fires (drafts have no message_id).
Updated website/docs/user-guide/messaging/telegram.md with a
'Streaming transport' section explaining auto|draft|edit|off, the
DM-only constraint, and the per-response fallback behaviour.
The existing _live_system_guard (PR #23397) blocked os.kill / os.killpg
and a narrow subset of subprocess invocations. Tests still SIGTERMed the
live gateway today (May 10) because the guard had structural holes.
Plug them all:
- subprocess: also wrap getoutput, getstatusoutput
- os.system, os.popen - completely unwrapped before
- pty.spawn - completely unwrapped before
- asyncio.create_subprocess_exec / create_subprocess_shell - bypassed
the subprocess module entirely; now wrapped
- Subprocess command inspection now looks at the WHOLE command string,
not just tokens[0]. Catches sudo systemctl, env systemctl, bash -c
'systemctl', setsid systemctl, /usr/bin/systemctl, etc.
- New process-killer block: pkill / killall / taskkill / fuser
targeting hermes/python patterns is now refused
- os.kill PID 0 (own group) allowed; PID -1 (every process we can
signal) refused
- subprocess.Popen wrapper preserves __class_getitem__ so third-party
packages that use Popen[bytes] as a type annotation still import
Coverage is locked in by tests/test_live_system_guard_self_test.py -
exercises every primitive against a guaranteed-foreign PID and asserts
the guard fires. Adding a new kill primitive without updating the guard
breaks CI.
scripts/run_tests.sh now also force-loads ~/.hermes/pytest_live_guard.py
when present (developer-machine convenience), so even worktrees that
predate this commit get the protection on subsequent test runs through
the canonical wrapper.
A Codex auxiliary timeout closes the underlying OpenAI client (so the
streaming hang doesn't sit until the user kills the session), but the
cached wrapper kept pointing at the now-dead transport. Subsequent
auxiliary calls (compression retry, memory flush, background review,
title generation routed via provider: main) reused that closed client
and failed fast with 'Connection error' until the gateway restarted —
even though the main agent route was healthy the whole time.
Sync `_get_cached_client` had no liveness check (async did, via loop
identity), and the connection-error fallback in `call_llm` only fired
on the auto provider path, so an explicit provider — including the
common `auxiliary.compression.provider: main` shape — never evicted.
Three fixes:
* New `_evict_cached_client_instance(target)` helper that drops the
cache entry whose stored client is target (or wraps it via
`_real_client`, for `CodexAuxiliaryClient`).
* `_CodexCompletionsAdapter._close_client_on_timeout` evicts the
wrapper after closing the inner OpenAI client.
* `call_llm` and `async_call_llm` evict on `_is_connection_error`
before re-raising, regardless of whether the provider is auto.
Net effect: one timeout costs one summary attempt + the existing 30s
compressor cooldown; the next compaction rebuilds the client and
works. Non-connection errors (4xx/5xx) do not evict, so cache hits
stay stable.
Closes#23432
Cherry-picked from PR #10371. Two-layer defense for the spurious-thread_id
issue (#3206):
1. _build_message_event filters DM thread_ids: only preserve thread_id
for real topic messages (is_topic_message=True). Telegram puts
message_thread_id on every DM that is a reply, but reply-chain ids
route to nonexistent threads on send.
2. _send_message_with_thread_fallback helper: control sends
(send_update_prompt, send_exec_approval / send_slash_confirm,
send_model_picker) retry once without message_thread_id when
Telegram returns BadRequest 'Message thread not found'. Mirrors
the pattern PR #3390 added for the streaming send path.
Salvage notes:
- Conflict 1 (line ~4099): merged the contributor's DM is_topic_message
filter with the existing forum General-topic default from #22423,
preserving both behaviors.
- Conflict 2 (line ~1664 / 1690): kept main's delete_message (PR #23416)
alongside the new helper. Tightened the helper's exception catch
from bare 'Exception' to use the existing _is_bad_request_error +
_is_thread_not_found_error helpers (line 484-496) for consistency
with the streaming send path.
- Widened the fix to send_update_prompt (was bare self._bot.send_message,
same bug class).
Authored by rahimsais via PR #10371 (re-attributed from donrhmexe@
local commit author).
* feat(goals): /goal checklist + /subgoal user controls
Two-phase judge for /goal — Phase A decomposes the goal into a detailed
checklist on first turn; Phase B evaluates each pending item harshly
against the agent's most recent response. The goal completes only when
every item is in a terminal status (completed or impossible). Adds
/subgoal so the user can append, complete, mark impossible, undo,
remove, or clear items the judge missed or got wrong.
Mechanics:
- GoalState gains `checklist` and `decomposed` fields, both backwards
compatible (old state_meta rows load unchanged).
- Phase A: aux call writes a harsh, exhaustive checklist; biased toward
more items not fewer. Falls through to legacy freeform judge when
decompose fails.
- Phase B: judge gets the checklist + last-response snippet + path to
a per-session conversation dump at <HERMES_HOME>/goals/<sid>.json.
A bounded read_file tool (max 5 calls per turn, restricted to that
one file) lets the judge inspect history when the snippet is
ambiguous. Stickiness in code: terminal items are frozen, only the
user can revert via /subgoal undo.
- Continuation prompt shows checklist progress when non-empty;
reverts to old prompt when empty.
- Status line shows M/N done counts.
CLI + gateway + TUI gateway all pass the agent reference into
evaluate_after_turn so the dump can be written. Gateway-side
/subgoal is allowed mid-run since it only modifies the checklist
the judge consults at turn boundaries.
Tests: 24 new cases — backcompat round-trip, Phase A decompose,
Phase B updates + new_items + stickiness, user override flows,
conversation dump (incl. unsafe-sid sanitization), judge read_file
restriction. Existing freeform-mode tests updated to patch the
renamed `judge_goal_freeform` and skip Phase A explicitly.
* fix(goals): off-by-one in judge index, message-list plumbing, prompt tuning
Three live-test findings from running /goal end-to-end against
gemini-3-flash-preview as the judge:
1. Off-by-one bug — the judge sees the checklist rendered with 1-based
indices ('1. [ ] foo, 2. [ ] bar') but the apply layer indexed
state.checklist as 0-based. Result: every judge update landed on
the wrong item, evidence got attached to neighbouring rows, and
the genuine 'first pending' item (usually #1) never got marked.
Fix: convert 1 → 0 in _parse_evaluate_response. Also tightened the
user prompt to call out the 1-based scheme explicitly. New tests
cover the parser conversion + an end-to-end fake-judge round-trip.
2. Conversation dump never happened — _extract_agent_messages tried
common AIAgent attribute names (.messages, .conversation_history,
etc.) but AIAgent doesn't expose the message list as an instance
attribute; it lives inside run_conversation()'s scope. Result: the
judge's read_file tool always saw history_path=unavailable. Fix:
added an explicit messages= kwarg to evaluate_after_turn that all
three call sites (CLI, gateway, TUI gateway) now pass directly.
Agent-attribute extraction kept as back-compat fallback.
3. Prompt was too harsh on simple goals. The original 'be HARSH,
default to leaving items pending' wording made the judge refuse
to mark 'file exists' completed even after the agent ran ls,
test -f, os.path.isfile, and find — burning the entire 8-turn
budget on a fizzbuzz task. Softened to 'strict but not absurd'
with explicit guidance on what counts as evidence and a directive
not to require re-proving items already established earlier.
Re-tested live with the same fizzbuzz goal: now terminates in 2
turns with all 8 checklist items correctly attributed to their
own evidence. /subgoal user-action flow (add / complete / undo /
impossible) verified live as well.
New TestUtf16OverflowDetection class covers two scenarios:
- test_emoji_text_exceeding_utf16_limit_triggers_overflow_split: feeds
2200 emoji codepoints (4400 UTF-16 units) — under Telegram's
codepoint-equivalent limit but over its UTF-16 limit. Asserts
truncate_message was called with len_fn=utf16_len, confirming the
consumer detected the overflow.
- test_codepoint_only_adapter_falls_back_to_len: documents that
adapters which don't subclass BasePlatformAdapter (or test MagicMocks)
fall back to plain len for backwards compat.
The contributor's PR shipped no tests for the UTF-16 path.
Slash commands (/clear, /new, /undo, /reload-mcp) are dispatched from the
process_loop daemon thread. prompt_toolkit.run_in_terminal returns a
coroutine that only the main-thread event loop can drive, so calling it
from a daemon thread orphans the coroutine — the input prompt never
renders and user keystrokes leak into the composer instead of the
confirmation prompt (issue #23185).
Mirror the thread-aware guard already in _run_curses_picker: when off the
main thread, fall back to a direct input() call. Also wrap
run_in_terminal in try/except so WSL / Warp / other emulators that
silently drop the scheduled coroutine fall back to input() too.
Tests: tests/cli/test_prompt_text_input_thread_safety.py covers main
thread (run_in_terminal path), daemon thread (direct input fallback),
no-app, run_in_terminal-raises, and EOF handling.
When kanban_complete rejects a created_cards list as hallucinated, the
task is intentionally left in-flight (the gate runs before the write
txn) so the worker can retry with a corrected list or pass
created_cards=[] to skip the check. The retry path already worked, but
the previous error wording read like a terminal failure and workers
were observed abandoning the run instead of trying again.
Spell out the recovery path explicitly in the tool_error response
("Your task is still in-flight ... Retry kanban_complete with ...") and
add regression coverage at both the kernel and tool layers so the
retry contract — and the wording the worker depends on to discover
it — is pinned.
Fixes#22923
Workers running slow models (e.g. kimi-k2.6) can spend longer than
DEFAULT_CLAIM_TTL_SECONDS inside a single tool-free LLM call, making
no tool calls and therefore not heartbeating. release_stale_claims
previously reclaimed these healthy workers, producing the
spawn-then-immediately-reclaim loop reported in #23025.
When a stale-by-TTL claim's host-local worker PID is still alive,
extend the claim (emit a claim_extended event) rather than killing
it. enforce_max_runtime / detect_crashed_workers remain the upper
bounds for genuinely wedged or dead workers. Reclaim events now also
record claim_expires, last_heartbeat_at, worker_pid, and host_local
so operators can see why a worker was killed.
xAI's Responses API returns HTTP 400 ("Model X does not support
parameter reasoningEffort") for grok-4, grok-4-0709, grok-4-fast-*,
grok-4-1-fast-*, grok-3, grok-4.20-0309-*, and grok-code-fast-1 — even
though those models reason natively. Hermes was unconditionally sending
`reasoning: {effort: 'medium'}` to xAI for every Grok model, breaking
direct `--provider xai` for the entire grok-4 line.
Add a substring allowlist predicate (verified live against api.x.ai
2026-05-10) covering the only Grok families that accept the effort dial:
grok-3-mini*, grok-4.20-multi-agent*, grok-4.3*. The Responses transport
omits the `reasoning` key entirely for everything else while still
including `reasoning.encrypted_content` so we capture native reasoning
tokens.
Verified end-to-end: `hermes chat -q hi --provider xai --model grok-4-0709`
went from HTTP 400 to a successful reply.
The contributor's regression test for Feishu fallback thread routing
asserted on attributes specific to the real lark SDK builder
(call_args.body, body.receive_id). In test environments without the
lark SDK installed, the in-tree fallback (gateway/platforms/feishu.py
_build_create_message_request) returns a SimpleNamespace using
.request_body instead of .body, causing AttributeError.
Now reads via getattr fallback and also verifies receive_id_type is
'thread_id' (not 'chat_id') as a stronger contract check.
When the first streamed message exceeds the platform length limit and
gets split into chunks, _send_new_chunk was called with self._message_id
(which is None on first send), dropping thread routing entirely.
Fallback to self._initial_reply_to_id so overflow chunks land in the
correct topic/thread.
Also fix a fragile test assertion that could be silently skipped.
Cherry-picked from PR #13077 commits:
- 5500c7d8 fix(gateway): stream consumer first message drops thread context
- e84403b9 test(gateway): add regression tests for stream consumer thread routing
Fixes: Streaming first message drops thread/topic context in Feishu group
topics, Slack threads, Telegram forum topics. Adds initial_reply_to_id
ctor arg to GatewayStreamConsumer, threaded through _send_or_edit and
_send_new_chunk. Also fixes Feishu _send_raw_message fallback path
(reply -> create) to use receive_id_type='thread_id' so the new message
lands in the correct topic instead of the main channel.
Authored by hrygo via PR #13077 (re-attributed from the bot-authored
salvage commit on the original branch).
The split-overflow path in _send_or_edit (gateway/stream_consumer.py) was
copying the cumulative _already_sent flag into _final_response_sent on the
done frame. _already_sent goes True on any successful prior edit (tool
progress) or on fallback-mode promotion when an edit fails — neither
proves the *current* chunked send delivered the final answer.
When the chunked send actually fails (network error, flood control), the
consumer would wrongly claim 'final delivered' and the gateway's
independent fallback delivery in run.py would be suppressed. User saw
only tool-progress bubbles and never got the answer.
Now we track per-chunk success locally: _send_new_chunk returns the new
message_id on success or returns the passed-in reply_to unchanged on
failure. If at least one returned id differs, chunks_delivered = True;
otherwise stays False, gateway fallback runs.
Adds two regression tests:
- test_split_overflow_failed_send_does_not_mark_final_sent — primes
_already_sent=True, then makes every send fail; asserts
_final_response_sent stays False.
- test_split_overflow_partial_send_marks_final_sent — happy path,
asserts _final_response_sent goes True.
Note: the companion bug at the CancelledError handler (issue cited
lines 417-418) was already fixed by 3b5572ded on 2026-04-16.
Closes#10748
Follow-up to the previous commit's notifier behavior change. Two test fixes:
1. `tests/gateway/test_kanban_notifier.py` gains
`test_notifier_redelivers_same_kind_on_dispatch_cycle` — pins the new
contract directly: a task that crashes, gets reclaimed, and crashes
again notifies the user BOTH times. Before #21398 the second crash
silently dropped because the subscription was already deleted.
2. `tests/hermes_cli/test_kanban_notify.py::
test_notifier_unsubs_after_abnormal_events[gave_up|crashed|timed_out]`
is flipped. Those tests were added in the salvage of #22941 and
asserted the OLD behavior (subscription deleted after gave_up /
crashed / timed_out). They're now obsolete — the new contract is
"subscription survives a non-final terminal event so retries reach
the user." Updated docstring + asserts; the cursor-advance check is
added to confirm the dedup mechanism still works.
The `test_notifier_unsubs_after_completed_event` test stays untouched
because `completed` IS still a terminal event that triggers unsub
(the task hits `done` status, which is handled by the `task_terminal`
branch in the notifier loop).
Follow-up to HuangYuChuh's #17384 cherry-pick:
- Use defensive getattr+logger.debug for delete_message lookup, mirroring
the sibling _try_send_fresh_final cleanup pattern at L820+. Platforms
that don't implement delete_message no longer raise AttributeError; the
failure path now logs at debug for diagnosability instead of silently
swallowing.
- Add three regression tests in tests/gateway/test_stream_consumer.py:
- delete_message awaited on happy-path exit with stale id
- delete_message NOT awaited when no fallback chunks reached the user
- no crash on adapters that lack delete_message (spec-restricted mock)
The shutdown forensics added in #23285 caught tests/hermes_cli/ pytest
runs sending SIGTERM to the developer's live gateway 5+ times in 3
days. Root cause: when a single test forgets to mock os.kill or
find_gateway_pids, the real call leaks past the hermetic HERMES_HOME
isolation — find_gateway_pids' psutil scan walks the whole machine and
returns the live gateway PID, then the unmocked os.kill delivers the
signal.
Rather than audit and patch ~30 tests across cmd_update, kill_gateway_processes,
and stop_profile_gateway code paths, install a single autouse guard in
tests/conftest.py that blocks the two primitives that actually cause
the damage:
- os.kill rejects any PID outside the test process subtree with a
hard RuntimeError so the offending test gets a stack trace instead
of silently murdering the real gateway.
- subprocess.run / Popen / call / check_call / check_output reject
any 'systemctl <verb> hermes-gateway' invocation that would mutate
the live unit. Read-only systemctl calls (status, show, list-units)
still pass through.
We intentionally do NOT stub find_gateway_pids / _scan_gateway_pids —
tests of those functions themselves need the real implementation.
Discovery without delivery is harmless; the os.kill + systemctl guards
catch the actual damage path.
Tests that legitimately need real signal delivery (e.g. PTY tests
signalling their own child) opt out via
@pytest.mark.live_system_guard_bypass.
Validation: tests/hermes_cli/ + tests/cli/ + tests/gateway/ produce
the same 17 failures with and without this guard (all pre-existing on
main, unrelated to gateway-kill leaks). The live gateway survives the
test run that previously SIGTERMed it.
Two follow-up improvements to the previous commit's notifier dedup work.
1. Add a regression test for the send-exception rewind path. The
contributor's PR included a test for the adapter-disconnect path
(test_kanban_notifier_rewinds_claim_if_adapter_disconnects, where
adapter is None at delivery time), but not for the "adapter is
connected, send() raises" path that fires inside the inner try/except
at gateway/run.py:4314. The new test
(test_kanban_notifier_rewinds_claim_on_send_exception) uses a
FailingAdapter that always raises and confirms (a) send was actually
attempted, (b) the claim was rewound, (c) the next call to
unseen_events_for_sub still returns the event for retry.
2. Drop the per-delivery success log from INFO to DEBUG. A busy board
on a multi-platform gateway can produce hundreds of these per day;
that's gateway.log noise that obscures real warnings. Failure paths
stay at WARNING (where you'd want to look when something's wrong)
so we don't lose visibility into transient send issues.
Builds on @kshitijk4poor's CLI handoff stub. The original PR's flow
deferred everything to whenever a real user happened to message the
target platform; this rewrites it so the gateway picks up handoffs
immediately and the destination chat just starts working.
State machine on sessions table replaces the boolean flag:
None -> 'pending' -> 'running' -> ('completed' | 'failed')
plus handoff_error for failure reasons. CLI request_handoff /
get_handoff_state / list_pending_handoffs / claim_handoff /
complete_handoff / fail_handoff helpers wrap the transitions.
CLI side (cli.py): /handoff <platform> validates the platform's home
channel via load_gateway_config, refuses if the agent is mid-turn,
flips the row to 'pending', and poll-blocks (60s) on terminal state.
On 'completed' it prints the /resume hint and exits the CLI like
/quit. On 'failed' or timeout it surfaces the reason and the CLI
session stays intact.
Gateway side (gateway/run.py): new _handoff_watcher background task
scans state.db every 2s, atomically claims pending rows, and runs
_process_handoff for each. _process_handoff:
1. Resolves the platform's home channel.
2. Asks the adapter for a fresh thread via the new
create_handoff_thread(parent_chat_id, name) capability so the
handed-off conversation gets its own scrollback. Adapters that
don't support threads (or fail) return None and the watcher
falls back to the home channel directly.
3. Constructs a SessionSource keyed as 'thread' when a thread was
created, 'dm' otherwise, then session_store.switch_session
re-binds the destination key to the CLI session_id. The full
role-aware transcript replays via load_transcript on the next
turn (no flat-text injection into context_prompt).
4. Forges a synthetic MessageEvent(internal=True) with the handoff
notice and dispatches through _handle_message; the agent runs
against the loaded transcript and adapter.send delivers the
reply.
5. Marks the row 'completed' on success, 'failed' (+error) on any
exception.
Adapter capability (gateway/platforms/base.py): create_handoff_thread
default returns None. Three overrides:
- Telegram (gateway/platforms/telegram.py): wraps _create_dm_topic
so DM topics (Bot API 9.4+) and forum supergroups both work.
- Discord (gateway/platforms/discord.py): parent.create_thread on
text channels with a seed-message + message.create_thread
fallback for permission edge cases. Skips DMs and other
non-thread-capable parents.
- Slack (gateway/platforms/slack.py): posts a seed message and
returns its ts as the thread anchor — Slack threads are
message-anchored.
In thread mode, build_session_key keys the destination without
user_id (thread_sessions_per_user defaults to False) so the synthetic
turn and any later real-user message in the thread share the same
session_key — seamless takeover without race.
CommandDef stays cli_only=True (handoff is initiated from the CLI;
gateway exposes /resume for the reverse direction).
Removed the original PR's _handle_message_with_agent handoff hook
(transcript-as-text injection into context_prompt) and the
send_message_tool notification — both replaced by the watcher path.
Tests rewritten around the new state machine: 13/13 pass.
E2E-validated thread + no-thread paths and the failure path against
real worktree imports with mocked adapters.
Adds /handoff <platform> CLI command that queues the current session for
resume on the configured home channel of any messaging platform.
CLI side:
- /handoff telegram — marks session in shared DB, sends summary to
the Telegram home channel via send_message
- /handoff discord — same for Discord
- Supports telegram, discord, slack, whatsapp, signal, matrix
Gateway side:
- On new session creation, checks for pending handoffs for the
incoming message's platform
- If found, loads the CLI session's full conversation history and
injects it into the context prompt as a handoff transcript
- Agent continues the conversation seamlessly
Files:
- hermes_state.py: handoff_pending, handoff_platform columns + helpers
- cli.py: _handle_handoff_command dispatch + handler
- hermes_cli/commands.py: CommandDef entry
- gateway/run.py: handoff detection in _handle_message_with_agent
- tests/hermes_cli/test_session_handoff.py: 8 tests
* feat(gateway): per-platform admin/user split for slash commands
Adds an opt-in two-list access control on top of the existing per-platform
`allow_from` allowlists, scoped to slash commands only:
- allow_admin_from — full slash command access
- user_allowed_commands — what non-admins may run
- group_allow_admin_from — same, group/channel scope
- group_user_allowed_commands
When `allow_admin_from` is unset for a scope, gating is disabled and every
allowed user keeps full access (backward compat). Plain chat is unaffected.
`/help` and `/whoami` are always reachable so users can see what they
can run.
Gate runs at the slash command dispatch site in gateway/run.py and uses
`is_gateway_known_command()`, so it covers built-in AND plugin-registered
commands through the live registry without per-feature wiring.
Adds `/whoami` showing platform, scope, tier, and runnable commands.
Salvage of PR #4443's permission tier work, scoped down. The full tier
system, tool filtering, audit log, usage tracking, rate limiting,
`/promote` flow, and persistent SQLite stores are not included here —
those can be re-expanded later if needed.
Co-authored-by: ReqX <mike@grossmann.at>
* fix(gateway): close running-agent fast-path bypass + add coverage and central docs
The slash command access gate was only applied at the cold dispatch site
(line ~5921). When an agent was already running, the running-agent
fast-path block (line ~5574) dispatched /restart, /stop, /new, /steer,
/model, /approve, /deny, /agents, /background, /kanban, /goal, /yolo,
/verbose, /footer, /help, /commands, /profile, /update directly
without going through the gate — letting non-admins bypass gating just
because an agent happens to be busy.
Refactored the gate into _check_slash_access() and called from BOTH
paths. /status remains intentionally pre-gate so users can always see
session state.
Also added 18 more dispatch tests covering:
- Running-agent fast-path: blocks non-admin, allows admin, /status
always works
- Alias canonicalization (gate uses canonical name, not user alias)
- Unknown / unregistered commands pass through (don't false-positive)
- DM admin scope-locked when group has its own admin list
- Multi-platform isolation (Discord gated, Telegram unrestricted)
Docs: added Slash Command Access Control section to the central
messaging index page + /whoami row in the chat commands table.
Co-authored-by: ReqX <mike@grossmann.at>
---------
Co-authored-by: ReqX <mike@grossmann.at>
When the gateway received SIGTERM, the shutdown_signal_handler ran a
synchronous 'ps aux' (3s timeout) inside the asyncio event loop, then
asyncio.create_task(runner.stop()). On a busy host that ate 1-3s of
the teardown budget before draining could even start, and the resulting
log line was a multi-line ps dump that didn't tell us who sent the
signal. The shutdown path itself logged 'Stopping gateway...' and then
nothing until 'Gateway stopped' — when systemd SIGKILLed mid-drain,
there was no way to see which phase wedged.
Changes:
- New gateway/shutdown_forensics.py:
* snapshot_shutdown_context(sig) — sub-millisecond /proc-only capture
of signal name, parent pid+name+cmdline, INVOCATION_ID (systemd
marker), loadavg_1m, TracerPid, takeover/planned-stop marker
presence + whether-it-names-self. Pure stdlib, never raises.
* spawn_async_diagnostic(log_path, sig) — detached subprocess with
its own 'timeout 5s', start_new_session=True, writes ps auxf +
pstree + dmesg to ~/.hermes/logs/gateway-shutdown-diag.log.
Returns immediately, can't block the event loop or the cgroup
teardown.
* check_systemd_timing_alignment(drain_timeout) — reads
/proc/self/cgroup for our unit, asks systemctl show for
TimeoutStopUSec, returns mismatch info when the unit's stop
timeout is smaller than restart_drain_timeout + 30s headroom
(the case where systemd SIGKILLs mid-drain).
* _parse_systemd_duration_to_us — covers '90s', '1min 30s',
'500ms', '1h' style values from systemctl show.
* format_context_for_log — single scannable key=value line, parent
cmdline last.
- gateway/run.py shutdown_signal_handler:
* Replaces synchronous ps aux + ad-hoc 'hermes-related lines' filter
with snapshot + detached spawn.
* Always logs 'Shutdown context: signal=... parent_pid=...
parent_cmdline=...' regardless of planned/unexpected so we can
correlate signal source even on planned restarts.
- gateway/run.py _stop_impl:
* Per-phase '+X.XXs' timing for notify_active_sessions, drain
(with drain_seconds, active_at_start, active_now, timed_out),
post-interrupt tool kill, each adapter disconnect (Xs),
all adapters disconnected, final-cleanup tool kill, SessionDB
close, total teardown.
- gateway/run.py start():
* Stale-unit warning at startup when the running systemd unit's
TimeoutStopSec is smaller than the configured drain timeout.
Points the user at 'hermes gateway service install --replace'
to regenerate, or at shortening agent.restart_drain_timeout.
Tests: 30 new in tests/gateway/test_shutdown_forensics.py — snapshot
speed bound, signal name resolution, marker detection self-vs-other,
async diag spawn doesn't block caller, systemd duration parser, and
alignment check returns None outside systemd. Wider tests/gateway/
suite: 5258 passing, 3 pre-existing TTS-routing failures unchanged
on main.
Follow-up to the previous commit's toolset-vs-skill validation.
The contributor's fix raises ValueError on the first toolset name found
in the skills list. That works for one mistake, but agents that confuse
skills with toolsets usually pass several at once
(`skills=["web", "browser", "terminal"]`) — and serial-correcting one
per failure round-trip wastes tokens. Collect all toolset-shaped
entries first, then raise once with the full list.
The error message is also slightly clearer:
'web', 'browser', 'terminal' are toolset names, not skill name(s).
Put toolsets in the assignee profile's `toolsets:` config instead of
per-task skills. Skills are named skill bundles (e.g. `kanban-worker`,
`blogwatcher`); toolsets are runtime capabilities (e.g. `web`,
`browser`, `terminal`).
vs. the previous "the assignee profile's toolsets" — explicitly naming
the YAML key (`toolsets:`) and giving concrete examples in both
categories closes the conceptual gap that produced the bug to begin
with.
Adds one regression test (test_create_task_skills_lists_all_toolset_typos)
covering the multi-name aggregation path. The single-typo test from
the original PR still passes (the loose `match="toolset name"` matches
both singular and plural forms).
Two follow-up improvements to Tranquil-Flow's metadata-panel restyle.
Both stay within the parent PR's "tone down the panel" scope.
1. Native <details>/<summary> collapse for verbose metadata.
The parent PR consciously deferred this ("adding native expand/collapse
would be the next step but requires UX agreement"). The default they
asked for is straightforward: collapsed when the rendered JSON exceeds
300 chars (the threshold where the max-height: 8.5rem cap actually
starts mattering), expanded otherwise. <details>/<summary> is the right
primitive — zero JS, browser-handled state, accessible by default
(keyboard-navigable, screen-reader announces the disclosure state),
and survives any react-state churn for free.
The OS-default disclosure marker is suppressed (list-style: none +
::-webkit-details-marker hidden) and replaced with a CSS ::before
chevron that rotates 90deg on the [open] attribute, so the look is
consistent across Firefox/WebKit/Blink without the double-marker
that would otherwise appear on the platforms that still render the
default triangle.
2. Skip rendering when metadata is an empty object.
`r.metadata && ...` truthy-checks, but `{}` is truthy in JS — so a
completed task with no actual metadata would render a "Metadata"
labeled disclosure block containing literal `{}`. Adds an
Object.keys(r.metadata).length > 0 guard so empty payloads render
nothing instead of an empty disclosure stub.
Tests: three new static-asset assertions covering the <details> shape,
the empty-object skip, and the suppress-default-marker + animated-chevron
CSS — all in `tests/plugins/test_kanban_dashboard_plugin.py`.
Hand-rebased onto current main from PR #19980; the original branch was stale
against main (~6 unrelated dashboard fixes had landed since), so applying
the PR's dist files directly would have silently reverted them.
The run-history panel in the task drawer rendered each completed run's
`metadata` field as a `<code class="hermes-kanban-run-meta">` containing
`JSON.stringify(r.metadata)` — a single unindented monoline. With
`white-space: pre-wrap` and a monospace font, a writer task's metadata
(changed_files paths, source URLs, generated-artifact details) wrapped
into a tall block of code-ish text that filled the parent run row. The
container's faint `--color-foreground 3%` background then made the whole
thing read like a crash dump even though the run completed normally.
Restyle and label, no interactivity changes:
- Wrap the meta payload in a `.hermes-kanban-run-meta-block` sub-block
with an explicit `Metadata` label (small, uppercase, muted) so the
panel reads as auxiliary detail at a glance.
- Pretty-print the JSON (`indent=2`) so the structure is scannable
instead of a wall of monoline text.
- Cap `.hermes-kanban-run-meta` at `max-height: 8.5rem; overflow: auto`
so a verbose blob scrolls inside its own pane rather than swamping
the run row.
- Sub-block uses a thin `border-left` rule and `background: transparent`
— distinct from the destructive-tinted treatment used by crashed /
timed_out / blocked / spawn_failed runs higher in the same file.
Tests: two new static-asset assertions in
`tests/plugins/test_kanban_dashboard_plugin.py` lock in the rendered
shape (the plugin ships built-only, no src/).
Adds CDPSupervisor.evaluate_runtime() and wires it into _browser_eval as a
fast path when a supervisor is alive for the current task_id. Replaces the
~180ms agent-browser subprocess fork+exec+Node-startup hop with a ~1ms
Runtime.evaluate over the supervisor's already-connected WebSocket.
Falls through to the existing agent-browser CLI path when no supervisor is
running (e.g. backends without CDP, or before the first browser_navigate
attaches one), so behaviour is unchanged where it can't apply.
JS-side exceptions surface directly without falling through to the
subprocess (the subprocess would just re-raise the same error, slower);
supervisor-side failures (loop down, no session) fall through cleanly.
Benchmark — 30 iterations of `1 + 1` against headless Chrome:
supervisor WS mean= 0.96ms median= 0.91ms
agent-browser subprocess mean=179.35ms median=167.73ms
→ 187x speedup mean
Tests: 14 unit tests (mocked supervisor + response-shape coverage), 5
real-Chrome e2e tests in test_browser_supervisor.py (gated on Chrome
being installed). Browser test suite: 355 passed, 1 skipped.
Follow-up to the previous commit's casing fix.
The original PR shipped the dist edits without test coverage. The
contributor's reasoning (UI-only attributes in a pre-built JS bundle,
nothing meaningful to unit-test) is fair, but a static-asset assertion
catches the most likely regression vector — a future rebuild of the
dist bundle that loses the attributes — at near-zero cost.
Adds two regression tests in tests/plugins/test_kanban_dashboard_plugin.py:
- test_dashboard_assignee_inputs_preserve_casing — reads dist/index.js
and asserts autoCapitalize="none", autoCorrect="off", spellCheck=false,
and textTransform="none" each appear at least twice (one per assignee
input — inline triage/lane create + task-edit panel).
- test_dashboard_lane_head_preserves_assignee_casing — reads dist/style.css
and asserts the .hermes-kanban-lane-head rule body does NOT contain
text-transform: uppercase. Locates the rule by marker so unrelated CSS
churn nearby doesn't flake the test.
Both follow the same shape as the existing test_dashboard_requests_default_board_explicitly
static-asset guard from PR #22940's salvage.
Also adds the AUTHOR_MAP entry for princepal9120's GitHub-noreply email
so release notes credit the right account.
Follow-up to the previous commit's safe-int task_age fix.
The original PR shipped without test coverage. This commit adds:
- test_safe_int_accepts_int_and_int_string — sanity for the well-typed
path so the helper itself can't quietly start swallowing valid values.
- test_safe_int_returns_none_on_corrupt_inputs — the failure modes
(None, '%s', 'abc', '', '1.5', random objects). Covers both the
ValueError and TypeError catch branches.
- test_task_age_handles_corrupt_created_at — the headline regression:
a task with created_at='%s' used to raise ValueError and turn
GET /api/plugins/kanban/board into a 500.
- test_task_age_handles_corrupt_started_and_completed — confirms the
safe-int treatment is consistent across all three timestamp fields.
- test_task_age_well_formed_task — regression that the safe path
doesn't change observable output for normal data.
- test_task_dict_survives_corrupt_created_at — defense in depth.
Writes a corrupt row directly via SQL, reads it back through the
ORM, and confirms task_age + the surrounding plugin_api guard
degrade gracefully instead of crashing.
Also adds the AUTHOR_MAP entry for the contributor's GitHub-noreply
email so release notes credit @baocin (the commit was authored locally
as `aoi <aoi@hino.local>` — re-attributed during salvage to the
github noreply form).
* feat(i18n): localize /model command output
Reported by @tianma8888: when Chinese users run /model, the labels
("Provider:", "Context:", "_session only_", etc.) are still English.
This routes the static prose through the existing i18n catalog so it
follows display.language / HERMES_LANGUAGE.
Changes:
- locales/{en,zh,ja,de,es,fr,tr,uk}.yaml: add 17 keys under
gateway.model.* covering switched/provider/context/max_output/cost/
capabilities/prompt_caching/warning/saved_global/session_only_hint/
current_label/current_tag/more_models_suffix/usage_*.
- gateway/run.py _handle_model_command: replace hardcoded f-strings in
the picker callback, the text-list fallback, and the direct-switch
confirmation block with t("gateway.model.<key>", ...).
What stays English:
- model IDs, provider slugs, capability strings, cost figures, and the
"[Note: model was just switched...]" prepended to the model's next
prompt (LLM-facing, not user-facing).
- The two slightly-different session-only hints unify on a single key
with the em-dash phrasing.
Validation: tests/agent/test_i18n.py 27/27 passing (parity contract
holds), tests/gateway/ -k 'model or i18n' 74/74 passing.
* feat(i18n): localize all gateway slash command outputs
Expands the i18n catalog from 7 strings to 234 keys across 35 gateway
slash command handlers, so non-English users see localized output for
\`/profile\`, \`/status\`, \`/help\`, \`/personality\`, \`/voice\`, \`/reset\`,
\`/agents\`, \`/restart\`, \`/commands\`, \`/goal\`, \`/retry\`, \`/undo\`,
\`/sethome\`, \`/title\`, \`/yolo\`, \`/background\`, \`/approve\`, \`/deny\`,
\`/insights\`, \`/debug\`, \`/rollback\`, \`/reasoning\`, \`/fast\`,
\`/verbose\`, \`/footer\`, \`/compress\`, \`/topic\`, \`/kanban\`,
\`/resume\`, \`/branch\`, \`/usage\`, \`/reload-mcp\`, \`/reload-skills\`,
\`/update\`, \`/stop\` (plus the \`/model\` block already added in the
previous commit).
Reported by @tianma8888 — Chinese users want command output prose in
their language, not just the labels we already had.
Translations are hand-written for all 8 supported locales (en, zh, ja,
de, es, fr, tr, uk), matching each catalog's existing style: full-width
punctuation in zh, em-dashes in zh/ja/uk, French spaced colons,
German noun capitalization, etc.
What stays English (unchanged):
- Identifiers/values: model IDs, file paths, profile names, session IDs,
command flag names like --global, URLs, config keys.
- Backtick code spans: \`/foo\`, \`config.yaml\`.
- Log messages (logger.info/warning/error).
- LLM-facing system notes prepended to next prompt (e.g. [Note: model
was just switched...]).
- Strings produced by external modules (gateway_help_lines,
format_gateway, manual_compression_feedback) — those have their
own surfaces.
New shared keys for cross-handler boilerplate:
- gateway.shared.session_db_unavailable (5 call sites: branch, title,
resume, topic, _disable_telegram_topic_mode_for_chat)
- gateway.shared.session_not_found (1 site)
- gateway.shared.warn_passthrough (2 sites in /title's f"⚠️ {e}" pattern)
YAML gotcha fixed: \`yolo.on\` and \`yolo.off\` were originally written
unquoted, which YAML 1.1 parses as boolean True/False keys. Renamed to
\`yolo.enabled\` / \`yolo.disabled\` for both safety and clarity.
Test fix: tests/agent/test_i18n.py::test_t_missing_key_in_non_english_falls_back_to_english
now resets the catalog cache on teardown, so the fake "foo: English Foo"
locale doesn't poison the module-level cache for subsequent tests in
the same xdist worker. (Without this, every gateway slash command test
that shares a worker with the i18n suite would see the fake catalog.)
Validation:
- tests/agent/test_i18n.py: 27/27 (parity contract — every key in every
locale, matching placeholder tokens).
- tests/gateway/: 5077 passed, 0 failed (full gateway suite).
- 180 t() call sites added across 35 handlers; 1872 catalog entries
total (234 keys × 8 locales).
* feat(i18n): add 8 new locales — af, ko, it, ga, zh-hant, pt, ru, hu
Expands the static-message catalog from 8 → 16 languages, each with full
270-key parity against the English source-of-truth. Every locale now
covers the same surface PR #22914 added: approval prompts plus all 35
gateway slash command outputs.
New locales:
- af Afrikaans (community ask in #21961 by @GodsBoy; PRs #21962, #21970)
- ko Korean (PRs #20297 by @tmdgusya, #22285 by @project820)
- it Italian (PR #20371 by @leprincep35700)
- ga Irish/Gaeilge (PR #20962 by @ryanmcc09-dot)
- zh-hant Traditional Chinese (PRs #20523 by @jackey8616, #13140 by @anomixer)
- pt Portuguese (PRs #20443 by @pedroborges, #15737 by @carloshenriquecarniatto, #22063 by @Magaav)
- ru Russian (PR #22770 by @DrMaks22)
- hu Hungarian (PR #22336 by @lunasec007)
Each locale uses native-quality translations matching the existing tone
and conventions of the older 8 locales:
- zh-hant uses 繁體 characters with TW/HK technical vocabulary (軟體
not 软件, 連線 not 连接, 設定 not 设置, 訊息 not 消息, 工作階段 not 会话, 程式
not 程序, 預設 not 默认, 伺服器 not 服务器), full-width punctuation 「：（）」.
- ko uses formal 합니다체 (습니다/합니다) register throughout.
- pt uses European Portuguese as baseline with neutral PT/BR vocabulary
where possible.
- ga uses standard An Caighdeán Oifigiúil; English loanwords retained
for tech terms without good Irish equivalents (gateway, API, JSON).
- All preserve {placeholder} tokens, backtick code spans, slash commands,
brand names (Hermes, MCP, TTS, YOLO, OpenAI, Telegram, etc.), and emoji.
Aliases added in agent/i18n.py:
- af-za, Afrikaans → af
- ko-kr, Korean, 한국어 → ko
- it-it, italiano → it
- ga-ie, Irish, Gaeilge → ga
- zh-tw, zh-hk, zh-mo, traditional-chinese → zh-hant (note: zh-tw used to
alias to zh; now aliases to its own zh-hant catalog)
- zh-cn, zh-hans, zh-sg → zh (unchanged from before)
- pt-pt, pt-br, brazilian, portuguese → pt
- ru-ru, Russian, русский → ru
- hu-hu, Magyar → hu
The zh-tw alias re-routing is intentional: previously typing 'zh-TW' got
the Simplified Chinese catalog (wrong vocabulary for Taiwan/HK users).
Now those users get the proper Traditional Chinese catalog.
Validation:
- tests/agent/test_i18n.py: 43/43 (parity contract holds for all 16
languages × 270 keys = 4320 catalog entries, with matching placeholder
tokens).
- E2E alias resolution verified for all 19 alias inputs (Afrikaans, ko-KR,
한국어, italiano, Gaeilge, zh-TW, zh-HK, traditional-chinese, pt-BR,
brazilian, Magyar, etc.).
- tests/gateway/: 5198 passed (3 pre-existing TTS routing failures
unrelated to i18n).
Credit to all contributors whose PRs surfaced these language requests.
Their original PRs may now be closed as superseded with credit.
* feat(dashboard-i18n): add 14 web dashboard locales matching the static catalog
Brings the React dashboard (web/src/) up to the same 16-language
coverage the static catalog already has after the previous commits in
this PR. The Translations interface is TypeScript-typed, so every new
locale must provide every key — tsc -b is the parity guard.
Languages added (each is a complete 429-line locale file):
- af Afrikaans
- ja Japanese (PR #22513 by @snuffxxx surfaced this)
- de German (PR #21749 by @mag1art)
- es Spanish (PR #21749)
- fr French (PRs #21749, #10310 by @foXaCe)
- tr Turkish
- uk Ukrainian
- ko Korean (PRs #21749, #18894 by @ovstng, #22285 by @project820)
- it Italian
- ga Irish (Gaeilge)
- zh-hant Traditional Chinese (PR #13140 by @anomixer)
- pt Portuguese (PRs #22063 by @Magaav, #22182 by @wesleysimplicio, #15737 by @carloshenriquecarniatto)
- ru Russian (PRs #21749, #22770 by @DrMaks22)
- hu Hungarian (PR #22336 by @lunasec007)
Each translation covers all 15 namespaces with full key parity vs en.ts,
preserves every {placeholder} token verbatim, keeps identifiers
untranslated (brand names, file paths, cron expressions, code spans),
translates the language.switchTo tooltip into the target language, and
matches existing tone conventions (zh-hant uses TW/HK vocab; ja uses
formal desu/masu; ko uses formal seumnida register; ga uses An
Caighdean Oifigiuil with English loanwords for tech vocab without good
Irish equivalents).
Plumbing:
- web/src/i18n/types.ts: Locale union expanded to all 16 codes.
- web/src/i18n/context.tsx: imports all 16 catalogs; exports
LOCALE_META (endonym + flag per locale); isLocale() type guard.
- web/src/i18n/index.ts: re-export LOCALE_META.
- web/src/components/LanguageSwitcher.tsx: replaced two-state EN-ZH
toggle with a click-to-open dropdown listing all 16 languages.
Note: zh-hant.ts exports zhHant (camelCase) since hyphen is invalid in
a JS identifier; the canonical 'zh-hant' string keys it in TRANSLATIONS.
Validation:
- npx tsc -b: 0 errors. Every locale satisfies Translations.
- npm run build (tsc + vite production): green, 2062 modules.
- Each locale file is exactly 429 lines.
Out of scope: plugin dashboards (kanban/achievements ship as prebuilt
bundles with no source in repo); Docusaurus docs (separate surface);
TUI (no i18n yet).
* feat(plugin-i18n): localize achievements + kanban plugin dashboards across all 16 locales
Brings the two shipped plugin dashboards (hermes-achievements, kanban)
under the same i18n umbrella as the core dashboard PR #22914 just
established. Both bundles now read user-facing strings from the host's
i18n catalog via SDK.useI18n() instead of hardcoded English.
## Approach
Plugin dashboards ship as prebuilt IIFE bundles in
plugins/<name>/dashboard/dist/index.js — no build step, no source in
repo (upstream-authored, vendored as compiled JS). Earlier contributor
PRs (#22594, #22595, #18747) tried direct edits but didn't actually
wire the bundles to read translations.
This change does the wiring properly:
1. Each bundle gets a useI18n shim at IIFE scope:
const useI18n = SDK.useI18n
|| function () { return { t: { kanban: null }, locale: "en" }; };
Older host SDKs without useI18n still load the bundle and render
English fallbacks.
2. A small tx(t, path, fallback, vars) helper resolves dotted keys
under the plugin's namespace (t.kanban.* or t.achievements.*) and
interpolates {placeholder} tokens.
3. Every React component starts with const { t } = useI18n() and
each user-visible string is wrapped in tx(t, "key", "English fallback").
Helpers called outside React components (window.prompt callers,
constants used during init) take t as a parameter.
4. Top-level constants that were English dictionaries (COLUMN_LABEL,
COLUMN_HELP, DESTRUCTIVE_TRANSITIONS, DIAGNOSTIC_EVENT_LABELS in
kanban) become getColumnLabel(t, status)-style functions backed by
FALLBACK_* dictionaries.
## Translations added
Two new top-level namespaces added to the dashboard's TypeScript-typed
Translations interface:
- achievements: ~70 keys covering the hero, scan banner, achievement
card, share dialog, stats, filters, and empty states.
- kanban: ~145 keys covering the board, columns (with nested
columnLabels and columnHelp sub-dicts), card detail panel,
bulk-actions toolbar, dependency editor, board switcher, and
diagnostic callouts.
Each key is provided across all 16 supported locales:
en, zh, zh-hant, ja, de, es, fr, tr, uk, af, ko, it, ga, pt, ru, hu.
Total new translation entries: ~3,440 (215 keys × 16 locales).
## What stays English (deliberate)
- API paths, CSS class names, data-* attributes, JSON keys, regex
strings, URLs, file paths (~/.hermes/kanban.db, boards/_archived/).
- State identifier strings used as lookup keys (triage / todo / ready /
running / blocked / done / archived) — labels translate, key strings
don't.
- The PNG share-card text rendered to canvas in the achievements
ShareDialog (HERMES AGENT watermark, UNLOCKED stamp, tier names) —
these become part of a globally-shared image and stay English.
- localStorage keys (hermes.kanban.selectedBoard).
- Brand names (Kanban, Hermes, WebSocket, Nous Research).
## Contributor credit
PR #22594 by @02356abc and PR #22595 by @02356abc supplied the
en + zh kanban namespace skeleton (145 keys); used as the en source-
of-truth in this commit and translated to the other 14 locales.
PR #18747 by @laolaoshiren first surfaced the achievements
localization request.
## Validation
- npx tsc -b: 0 errors. All 16 locale .ts files satisfy the
Translations type with full key parity.
- npm run build (tsc + vite production build): green, 2062 modules,
1.56MB JS / 95KB CSS, ~2.5s build.
- node --check on both plugin bundles: parse cleanly.
- 126 tx() call sites in kanban, 46 in achievements.
## Out of scope
- TUI (ui-tui/) has no i18n infrastructure yet.
- Docusaurus docs (website/i18n/) — already had zh-Hans; expanding
is a separate translation workstream (Thai / Korean / Hindi PRs).
Follow-up to the previous commit's contributor cherry-pick.
The cherry-picked change replaced the bare ``["hermes", ...]`` spawn with
``[sys.executable, "-m", "hermes", ...]``. The intent was right (avoid
PATH dependence — cron, systemd User= services, launchd jobs, and other
detached dispatcher invocations routinely run with a stripped $PATH that
doesn't include the venv's bin/, breaking the bare-shim spawn) but the
module name is wrong: there is no top-level ``hermes`` package. The
console-script entry point in pyproject.toml is
``hermes = "hermes_cli.main:main"``, and ``python -m hermes`` fails with
``No module named hermes``. The cherry-picked form would have replaced a
sometimes-broken spawn with an always-broken one.
This commit:
- Adds ``_resolve_hermes_argv()``, mirroring ``gateway.run._resolve_hermes_bin``.
Tries ``shutil.which("hermes")`` first (preferred — keeps existing ``ps``
output and log lines familiar in the common case) and falls back to
``[sys.executable, "-m", "hermes_cli.main"]`` when the shim is not on
PATH. The fallback goes through the running interpreter so it's
PATH-independent. Kept as a local helper rather than imported from
gateway because ``hermes_cli`` sits below ``gateway`` in the dependency
order.
- Switches the dispatcher's ``cmd`` list to use ``*_resolve_hermes_argv()``.
- Adds three regression tests:
* ``test_resolve_hermes_argv_prefers_path_shim`` — pins the PATH-first
branch so a future refactor doesn't silently flip the order.
* ``test_resolve_hermes_argv_falls_back_to_module_form_when_no_path_shim`` —
pins the correct module name (``hermes_cli.main``, NOT ``hermes``).
Direct regression guard for the form that shipped in the original PR.
* ``test_resolve_hermes_argv_module_actually_runs`` — runs the fallback
invocation as a real subprocess and asserts ``--version`` works, so
losing ``hermes_cli.main``'s ``__main__`` handling can't slip past the
string-match test.
Verified end-to-end: with the shim on PATH the resolver returns
``[/.../hermes]`` and ``--version`` works; with the shim removed the
resolver returns ``[python, -m, hermes_cli.main]`` and ``--version``
still works; the original PR's ``python -m hermes`` invocation fails as
expected (``No module named hermes``).
* feat(plugins): host-owned LLM access via ctx.llm
Plugins can now ask the host to run a one-shot chat or structured
completion against the user's active model and auth, without ever
seeing an OAuth token or API key. Closes the gap where plugins that
needed bounded structured inference (receipts, CRM extraction,
support classification) had to either bring their own provider keys
or register a tool the agent had to call.
New surface on PluginContext:
- ctx.llm.complete(messages, ...)
- ctx.llm.complete_structured(instructions, input, json_schema, ...)
- async siblings ctx.llm.acomplete / acomplete_structured
Backed by the existing auxiliary_client.call_llm pipeline — every
provider, fallback chain, vision routing, and timeout policy Hermes
already supports applies automatically.
Trust gate (fail-closed by default):
- plugins.entries.<id>.llm.allow_model_override
- plugins.entries.<id>.llm.allowed_models (allowlist; '*' = any)
- plugins.entries.<id>.llm.allow_agent_id_override
- plugins.entries.<id>.llm.allow_profile_override
Embedded model@profile shorthand goes through the same gate as
explicit profile=, so it can't bypass the auth-profile policy.
Conflicting explicit and embedded profiles fail closed.
Also lands:
- plugins/plugin-llm-example/ — reference plugin that registers
/receipt-extract, demonstrating image+text structured input,
jsonschema validation, and the trust-gate config.
- website/docs/developer-guide/plugin-llm-access.md — full API docs.
- 45 unit tests covering trust gates, JSON parsing, schema
validation, image encoding, async surface, and config loading.
Validation:
- 2628 tests pass in tests/agent/
- E2E: bundled plugin loaded with isolated HERMES_HOME, slash
command produced parsed JSON via stubbed call_llm
- response_format extra_body wired correctly for both json_object
and json_schema modes
* docs(plugin-llm): rewrite quickstart and framing
The quickstart now uses a meeting-notes-to-tasks example instead of
a receipt extractor, and the page leads with hook-time / gateway
pre-filter / scheduled-job framing rather than the OpenClaw
KB/support/CRM/finance/migration enumeration that the original
upstream PR used. Receipt example moved to a separate worked
example link so the docs page itself doesn't echo any of the
upstream framing.
Also clarifies where ctx.llm fits in the broader plugin surface
(table comparing register_tool / register_platform / register_hook
/ etc.) and what makes this lane different from auxiliary_client
internals.
No code change.
* docs(plugin-llm): reframe as any LLM call, not just structured output
The original draft leaned heavily on complete_structured() and made
the chat lane (complete() / acomplete()) feel like a footnote.
Restructure so:
- The page title and description say 'any LLM call.'
- The lead shows BOTH a plain chat call (error rewriter) AND a
structured call (triage scorer) up top.
- Quick start has two complete plugin examples — /tldr (chat) and
/paste-to-tasks (structured).
- New 'When to use which' table for choosing complete() vs
complete_structured() vs the async siblings.
- Trust-gate sections explicitly note 'all four methods,' and the
request-shaping list calls out chat-only fields (messages) and
structured-only fields (instructions, input, json_schema)
alongside each other.
- The 'Where this fits' section now says 'for any reason,
structured or not.'
The receipt-extractor reference plugin still exists under
plugins/plugin-llm-example/ — but the docs page no longer treats
it as the canonical surface example. It's now described as 'a third
worked example, this time with image input.'
No code change.
* feat(plugin-llm): split provider/model into independent explicit kwargs
The first cut accepted a single 'provider/model' slug on every method
and split it internally. That looked clean but broke under live test:
the model-override path tried to use the slug's vendor prefix as a
literal Hermes provider id, which silently switched the user off
their aggregator (e.g. plugin asks for 'openai/gpt-4o-mini' on a user
who routes through OpenRouter — host attempted to call the 'openai'
provider directly, failed because OPENAI_API_KEY wasn't set).
New shape mirrors the host's main config:
ctx.llm.complete(
messages=[...],
provider='openrouter', # gated, optional
model='openai/gpt-4o-mini', # gated, optional
profile='work', # gated, optional
...
)
Each is independently gated by its own allow_*_override flag.
Granting model-override does NOT auto-grant provider-override.
Allowlists are now per-axis (allowed_providers, allowed_models)
matched literally against whatever string the plugin sends.
Dropped 'model@profile' embedded-suffix shorthand entirely. Hermes
doesn't use that pattern anywhere else; profile= is its own kwarg.
Live E2E (against real OpenRouter via Teknium's config) confirms:
- zero-config call works
- default-deny blocks each override with a helpful error
- model-only override stays on user's active provider (the bug)
- provider+model override switches cleanly
- allowlist refuses non-listed entries
- structured output round-trip parses + schema-validates
Tests: 49 cases (up from 45); all green. Docs updated to match the
new shape, including a 'most plugins never need this section' callout
on the trust-gate config block.
* fix+cleanup(plugin-llm): real attribution, hook-mode coverage, move example out of core
Three integration fixes for the ctx.llm surface:
1. Attribution bug — result.provider and result.model now reflect
what call_llm actually used, not placeholder fallbacks ('auto',
'default'). New _resolve_attribution() helper:
- explicit overrides win (what the call targeted)
- response.model wins for the recorded model (provider
canonicalisation: 'gpt-4o' → 'gpt-4o-2024-08-06' etc.)
- falls back to _read_main_provider() / _read_main_model()
when no override is set, so audit logs reflect the user's
active main provider/model
- 'auto' / 'default' only when EVERYTHING is empty
Live verified: zero-config call now records
provider='openrouter', model='anthropic/claude-4.7-opus-20260416'
instead of provider='auto', model='default'.
2. Hook-mode coverage — TestHookMode confirms ctx.llm.complete
works from inside a registered post_tool_call callback. The
docs page promised hook integration; now there's a test that
exercises the lazy-import path through the real invoke_hook
machinery. Two cases: traceback-rewrite hook with conditional
ctx.llm.complete, and minimal hook regression for the
sync-hook + sync-llm path.
3. Reference plugin moved out of core. plugins/plugin-llm-example/
is gone from hermes-agent — it now lives in the new
NousResearch/hermes-example-plugins companion repo. The docs
page links there. Hermes' bundled plugins should be plugins
users actually run; reference / docs-companion plugins live
externally.
Test count: 56 (up from 49). Wider sweep on tests/hermes_cli/
+ tests/gateway/ + tests/tools/ + tests/agent/ shows 16770
passing; the 12 failures are all pre-existing on origin/main
(verified by stashing this branch's changes and re-running) —
kanban-boards, delegate-task, gateway-restart, tts-routing —
none touch the plugin_llm surface.
* chore(plugins): move all example plugins to companion repo
Reference / docs-companion plugins now live exclusively in
NousResearch/hermes-example-plugins, not bundled with the core repo:
- example-dashboard
- strike-freedom-cockpit
A new fourth example, plugin-llm-async-example, was added to that
repo demonstrating ctx.llm's async surface (acomplete()) with
asyncio.gather() — registers /translate <lang>: <text> which fires
forward translation + sentiment classifier in parallel, then a
back-translation for QA. Live-tested at 2.5s for three real
provider round-trips (would be ~5-6s sequential).
Docs updated:
- developer-guide/plugin-llm-access.md links both sync and async
examples in the Reference section
- user-guide/features/extending-the-dashboard.md repoints both demo
sections to the companion repo with corrected install paths
- user-guide/features/built-in-plugins.md drops the two demo rows
- AGENTS.md notes that example plugins live in the companion repo
Net: hermes-agent's plugins/ directory now contains only plugins
users actually run (memory providers, dashboard tabs that ship real
features, the disk-cleanup hook, platform adapters). All four
demo / reference plugins live externally where they can be cloned
on demand instead of inflating the core install.
Follow-up to the previous commit's middleware fix.
- plugins/kanban/dashboard/plugin_api.py: rewrite the "Security note"
docstring. The previous text said "/api/plugins/ is unauthenticated by
design" — that's now actively wrong and dangerously misleading. New
text explains that plugin routes flow through the same session-token
middleware as core API routes and that --host 0.0.0.0 is safe to use
on a LAN as a result.
- tests/hermes_cli/test_web_server.py: extend TestPluginAPIAuth to cover
the surfaces the original PR didn't pin:
* test_plugin_route_allows_auth now exercises a real plugin path
(/api/plugins/example/hello) instead of accepting 200 OR 404 from
a maybe-loaded kanban plugin — the assertion was effectively vacuous.
* test_plugin_patch_requires_auth + test_plugin_delete_requires_auth
cover non-GET mutation methods in case a future regression
whitelists them by accident.
* test_non_kanban_plugin_route_requires_auth proves the fix is
plugin-agnostic, not kanban-specific (hits hermes-achievements +
a non-existent plugin namespace; both 401 before route resolution).
* test_plugin_websocket_unaffected_by_http_middleware locks in that
the HTTP middleware change didn't accidentally start gating WS
upgrades — kanban /events still uses its own ?token= check.
Plus a cosmetic blank-line cleanup.
Remove the blanket /api/plugins/* exemption from auth_middleware so
plugin API routes (e.g. Kanban dashboard) require the same session
token as all other /api/ endpoints.
Fixes#19533
Surfaces the pin command at the moment users care about it: when a
consolidation just landed against their skill library and they're
looking at the umbrella name in the curator output. Previously `hermes
curator pin` existed but had no discovery surface — users only learned
it existed by reading docs or stumbling onto `hermes curator --help`.
The hint:
archived 3 skill(s):
• docx-extraction → document-tools
• pdf-extraction → document-tools
• old-stale — pruned (stale)
full report: hermes curator status
keep an umbrella stable: hermes curator pin document-tools
Gated on having at least one consolidation that produced an umbrella.
Pruned-only runs (nothing surviving to pin) skip the hint. When
multiple umbrellas were produced, picks alphabetically first as a
concrete example rather than listing them all.
3 new tests in tests/agent/test_curator_classification.py covering:
consolidation produces hint with real umbrella name, pruned-only run
omits it, multi-umbrella picks one example.
* feat(gateway): add LINE Messaging API platform plugin
Adds LINE as a bundled platform plugin under `plugins/platforms/line/`,
synthesized from the strongest pieces of seven open community PRs. The
adapter requires zero core edits — `Platform("line")` is auto-discovered
via the bundled-plugin scan in `gateway/config.py`, and all hooks
(setup, env-enablement, cron delivery, standalone send) are wired
through `register_platform()` kwargs the way IRC and Teams do it.
Highlights merged into one plugin:
- **Reply token preferred, Push fallback.** Try the free reply token
first (single-use, ~60s TTL); fall back to metered Push when the
token is absent, expired, or rejected. (PR #21023)
- **Slow-LLM Template Buttons postback.** When the LLM is still running
past `LINE_SLOW_RESPONSE_THRESHOLD` (default 45s), the adapter burns
the original reply token to send a "Get answer" button bubble. The
user taps it to fetch the cached answer via a fresh reply token —
also free. State machine: PENDING → READY → DELIVERED, ERROR for
cancelled runs (orphan resolves to `LINE_INTERRUPTED_TEXT` after
/stop). Set threshold to 0 to disable. (PR #18153)
- **Three-allowlist gating** — separate user / group / room allowlists
with `LINE_ALLOW_ALL_USERS=true` dev-only escape hatch. (PR #18153)
- **Markdown URL preservation.** Strip bold/italic/code-fence/heading
markers (LINE renders them literally) but keep `[label](url)` →
`label (url)` so URLs stay tappable. (PR #18153)
- **System-message bypass** for `⚡ Interrupting`, `⏳ Queued`, etc. —
busy-acks reach the user as visible bubbles instead of being
swallowed into the postback cache. (PR #18153)
- **Media via public HTTPS URLs.** LINE doesn't accept binary uploads;
images/audio/video must be HTTPS-reachable. The adapter serves
registered tempfiles under `/line/media/<token>/<filename>` from the
same aiohttp app. Allowed-roots traversal guard covers
`tempfile.gettempdir()`, `/tmp` (→ `/private/tmp` on macOS), and
`HERMES_HOME`. `LINE_PUBLIC_URL` overrides URL construction for
setups behind tunnels/proxies. (PR #8398)
- **5-message-per-call batching.** LINE rejects >5 messages per
Reply/Push; smart-chunker caps text at 4500 chars per bubble.
- **Inbound dedup** via `webhookEventId` LRU. (PR #21023)
- **Self-message filter** via `/v2/bot/info` userId lookup. (PR #21023)
- **Loading-animation indicator** wired to LINE's `chat/loading/start`
endpoint, DM-only (LINE rejects it for groups/rooms). (PR #21023)
- **Out-of-process cron delivery** via `_standalone_send`, so
`deliver: line` cron jobs work even when cron runs detached from
the gateway.
- **Webhook hardening** — 1 MiB body cap, constant-time HMAC-SHA256
signature verification, dedup, scoped lock so two profiles can't
bind the same channel.
Validation
----------
- `scripts/run_tests.sh tests/gateway/test_line_plugin.py` →
73 passed in 1.05s
- `scripts/run_tests.sh tests/gateway/test_line_plugin.py
tests/gateway/test_irc_adapter.py
tests/gateway/test_plugin_platform_interface.py
tests/gateway/test_platform_registry.py
tests/gateway/test_config.py` → 193 passed, 7 skipped
- E2E import + register + signature roundtrip + `Platform("line")`
bundled-plugin discovery verified against current `origin/main`.
Closes the seven open LINE PRs (#18153, #16832, #6676, #21023, #14942,
#14988, #8398) by superseding them with a single plugin-form
implementation that takes the best idea from each.
Co-authored-by: pwlee <32443648+leepoweii@users.noreply.github.com>
Co-authored-by: Jetha Chan <jetha@google.com>
Co-authored-by: Cattia <openclaw@liyangchen.me>
Co-authored-by: perng <charles@perng.com>
Co-authored-by: Soichiro Yoshimura <soichiro0111.dev@gmail.com>
Co-authored-by: David Zhou <77736378+David-0x221Eight@users.noreply.github.com>
Co-authored-by: Yu-ga <74749461+yuga-hashimoto@users.noreply.github.com>
* docs(platforms): document platform-specific slow-LLM UX pattern
Add a 'Platform-Specific Slow-LLM UX' section to the platform-adapter
developer guide covering the _keep_typing override pattern that LINE
uses for its Template Buttons postback flow.
Three subsections:
- Pattern: subclass _keep_typing to layer mid-flight UX (with code)
- Pattern: subclass send to route through a cache instead of sending
- When this pattern is appropriate (vs. always-Push fallback)
Plus a short pointer in gateway/platforms/ADDING_A_PLATFORM.md so
tree-readers find the prose walkthrough on the docsite.
Filed because the LINE plugin (PR #23197) was the first bundled
adapter to need this pattern — every prior plugin (irc, teams,
google_chat) handles slow responses with the default typing-loop and
a regular send_text. Documenting now while the rationale is fresh.
---------
Co-authored-by: pwlee <32443648+leepoweii@users.noreply.github.com>
Co-authored-by: Jetha Chan <jetha@google.com>
Co-authored-by: Cattia <openclaw@liyangchen.me>
Co-authored-by: perng <charles@perng.com>
Co-authored-by: Soichiro Yoshimura <soichiro0111.dev@gmail.com>
Co-authored-by: David Zhou <77736378+David-0x221Eight@users.noreply.github.com>
Co-authored-by: Yu-ga <74749461+yuga-hashimoto@users.noreply.github.com>
Adapted from PR #20568 commit ce3518578 (Eric Litovsky / @kallidean).
Adds two-tier gating for the kanban tool surface so dispatcher-spawned
workers see only task-lifecycle tools (show/complete/block/heartbeat/
comment/create/link) while orchestrator profiles with `toolsets: [kanban]`
also see board-routing tools (kanban_list, kanban_unblock).
Workers shouldn't be enumerating or unblocking the board — they should
close their own task via the lifecycle tools. Hiding board-routing tools
from worker schemas keeps the worker focused and the toolset-isolation
contract honest.
Plus inherited from the same upstream commit:
- 50/200 row bound on kanban_list with `truncated` + `next_limit` metadata.
- Belt-and-suspenders runtime guard `_require_orchestrator_tool()` inside
the orchestrator handlers in case a stale registration ever routes a
worker to one of them.
- Tests for the new gate, the stricter bound, and the fact that even a
worker with `toolsets: [kanban]` in config still doesn't see board
routing.
Co-authored-by: Eric Litovsky <elitovsky@zenproject.net>
Two follow-ups from self-review:
1. Add gpt-5.3-codex-spark to DEFAULT_CONTEXT_LENGTHS at 128k. The
primary resolution path for Spark goes through provider='openai-codex'
→ _CODEX_OAUTH_CONTEXT_FALLBACK (already correct). But if any future
code path resolves Spark's context with a different provider (custom
proxy, generic fallthrough), the longest-substring-first lookup in
step 8 would match 'gpt-5' and report 400k, which is wrong by ~3x.
Adding the explicit override is a cheap defensive correctness fix
matching how gpt-5.4-mini and gpt-5.4-nano already shadow the generic
gpt-5 entry.
2. Update test_openai_codex_model_validation_fallback.py docstring. The
bug it was originally written for (gpt-5.3-codex-spark missing from
listing) is now resolved by this PR's catalog restoration. The test
still validly exercises the soft-accept code path for any future
entitlement-gated Codex slug that ships before Hermes catalogs it,
but the framing was stale — clarified.
Two follow-ups from self-review:
1. Add unit test for _fetch_models_from_api covering the live HTTP path.
The salvaged PR #19530 dropped the supported_in_api:false filter in
both _fetch_models_from_api and _read_cache_models, but only the
cache path had a regression test. This adds the symmetric live-fetch
test (mocked httpx) so a future drive-by change to the HTTP path
can't silently re-introduce the filter.
2. Pin test_codex_picker_uses_live_codex_catalog to the cache fallback.
The test wrote a fake JWT and a CODEX_HOME cache, but provider_model_ids
('openai-codex') still issued a real 10s HTTP probe to
chatgpt.com/backend-api/codex/models before falling back to the cache.
That made the test slow and non-deterministic in restricted/CI
networks. Patch _fetch_models_from_api to return [] so we go straight
to the cache path the test actually means to exercise.
Closes#6051.
Reported failure mode: agent migrated to WSL2, browser launch failed
because Playwright wasn't installed yet. Background reviewer captured
the failure as a durable skill (`browser-tool-launch-issue`) and the
agent kept refusing the browser tool for weeks after Playwright was
installed and verified working. Negative claims also propagated into
unrelated skills ("browser tools do not work", "cannot use Y from
execute_code").
Root cause: `_SKILL_REVIEW_PROMPT` and `_COMBINED_REVIEW_PROMPT` both
lean hard on "be active, save things, a pass that does nothing is a
missed learning opportunity." Neither distinguished durable knowledge
from transient environment state. The reviewer was doing what it was
told.
Fix at the write site — both prompts now carry a "Do NOT capture"
section calling out:
• Environment-dependent failures (missing binaries, fresh-install
errors, post-migration path mismatches, 'command not found',
unconfigured credentials, uninstalled packages)
• Negative claims about tools or features ("X does not work")
that harden into self-cited refusals
• Session-specific transient errors that resolved before the
conversation ended
• One-off task narratives ("summarize today's market", "analyze
this PR") — also addresses the #12812 / #4538 family
Plus a positive-reframing line: when a tool fails because of setup
state, capture the FIX (install command, config step, env var)
under an existing setup/troubleshooting skill — never "this tool
doesn't work" as a standalone constraint.
Targeted tests: 24/24 passing in tests/run_agent/test_review_prompt_class_first.py
(2 new + all existing review-prompt assertions). Substring-based
checks so future prompt edits don't false-fail.
The previous PR (#22993) gave us a structured WARNING per stream drop
but the only diagnostic was 'error_type=APIError error=Network
connection lost.' — same nothing the user started with. To actually
diagnose why subagents drop streams disproportionately we need to know
WHERE the drop happened.
Adds three breadcrumbs to the agent.log WARNING:
1. Inner exception chain. openai SDK wraps httpx errors as
APIConnectionError / APIError so the catch site only sees the
wrapper. _flatten_exception_chain walks __cause__/__context__ up to
4 levels deep and renders 'Outer(msg) <- Inner(msg)' so we can
tell ConnectError vs RemoteProtocolError vs ReadError vs
ProxyError without enabling verbose mode.
2. Upstream HTTP headers. Snapshots cf-ray, x-openrouter-provider,
x-openrouter-model, x-openrouter-id, x-request-id, server, via,
etc. from stream.response immediately after open (so they survive
even when the stream dies before the first chunk). These answer
'is one CF edge / one downstream provider responsible, or random?'
3. Per-attempt counters. bytes streamed, chunk count, elapsed time on
the dying attempt, and time-to-first-byte. Distinguishes 'couldn't
connect at all' (0s, 0 bytes) from 'died after 30s mid-stream'
(very different root causes — first is auth/routing, second is
upstream idle-kill or proxy timeout).
Plumbing:
- _stream_diag_init / _stream_diag_capture_response live on AIAgent
and produce a per-attempt dict held on request_client_holder['diag']
for closure access from the retry block.
- _call_chat_completions and _call_anthropic both initialize the diag
and increment counters per chunk/event (best-effort, never raises in
the streaming hot path).
- _log_stream_retry / _emit_stream_drop accept an optional diag and
render the new fields. Final-exhaustion log goes through the same
helper so it gets the same diagnostic dump.
- UI status line gains a brief 'after Xs' suffix when timing is
available — distinguishes 'connect failed' from 'died mid-stream'
at a glance without grepping logs.
Sample WARNING after this change:
Stream drop mid tool-call on attempt 2/3 — retrying.
subagent_id=sa-2-cafef00d depth=1 provider=openrouter
base_url=https://openrouter.ai/api/v1
error_type=APIError error=Connection error.
chain=APIError(Connection error.) <- RemoteProtocolError(peer
closed connection without sending complete message body)
http_status=200 bytes=12400 chunks=47 elapsed=12.00s ttfb=0.83s
upstream=[cf-ray=8f1a2b3c4d5e6f7g-LAX
x-openrouter-provider=Anthropic
x-openrouter-id=gen-abc123 server=cloudflare]
Tests: 10 covering diag init, header capture (whitelist enforced for
PII), exception-chain walking + depth cap, log content with full diag,
log content without diag (placeholders), UI elapsed-suffix on/off.
Closes#21794.
`/kanban`, `/kanban help`, `/kanban --help`, and `/kanban <sub> -h`
all returned broken output to the gateway and interactive CLI. Three
underlying bugs in `hermes_cli.kanban.run_slash`:
1. argparse writes help to **stdout** but `run_slash` only captured
stderr at parse time, so `-h` text was silently swallowed and
replaced with the `(usage error: 0)` sentinel.
2. The wrapping parser used `prog="/"` and routed via a synthetic
"_top → kanban" subparser, producing `usage: / kanban …` (stray
space) and `usage: /kanban kanban …` (doubled token) in error text.
3. Bare `/kanban` and `/kanban help` dumped argparse's full ~3KB
usage tree, which reads as visual garbage in a chat bubble.
Fix: drive the kanban_parser directly (no double-wrap), rewrite prog
strings on every leaf subparser, capture stdout AND stderr around
parse_args, distinguish SystemExit(0) (help — return captured stdout)
from SystemExit(2) (error — return single-line ⚠-prefixed message),
and add an explicit chat-friendly short-help block returned for bare
invocation and the help aliases (`help`, `--help`, `-h`, `?`).
Added 5 regression tests covering bare invocation, every help alias,
subcommand help, unknown action, and missing required arg.
Affects every chat platform via gateway/run.py::_handle_kanban_command
and the interactive CLI via cli.py::_handle_kanban_command.
Co-Authored-By: Nagatha (Claude Opus 4.7) <noreply@anthropic.com>
Both `_kanban_notifier_watcher` and `_kanban_dispatcher_watcher`'s
`_tick_once_for_board` called `_kb.connect(board=slug)` immediately
followed by `_kb.init_db(board=slug)`. Since `connect()` already runs
the schema + idempotent migration on first open per process, the
explicit `init_db()` was redundant — and worse, `init_db()` deliberately
busts the per-process `_INITIALIZED_PATHS` cache and re-runs the migration
on a *second* connection that races the first.
On every cold gateway start against a legacy DB this surfaced as either
`sqlite3.OperationalError: duplicate column name: <col>` or intermittent
`database is locked` errors logged at the first tick. The duplicate-column
case is now tolerated by `_add_column_if_missing` (commit 78698381a), but
the wasted second migration plus the database-is-locked race remain
fixable by skipping the redundant call entirely.
Drops `_kb.init_db(board=slug)` at both call sites and adds a regression
test in `tests/hermes_cli/test_kanban_notify.py` that pins the absence
via source inspection plus a runtime spy.
Co-authored-by: Teknium <127238744+teknium1@users.noreply.github.com>
Subagent stream drops were spamming the parent terminal with two lines
per blip ('Connection dropped...' + 'Reconnected...') while leaving zero
breadcrumb in agent.log to debug them.
Two underlying bugs, fixed together:
1. quiet_mode raised the run_agent/tools/etc. loggers to ERROR, which
filters records before root-logger file handlers see them. The comment
claimed 'File handlers still capture everything' — that was wrong.
Removed in both run_agent.py and cli.py; console quietness already
comes from hermes_logging not installing a console StreamHandler in
non-verbose mode.
2. The stream-retry blocks emitted two _emit_status calls per drop
('⚠️ Connection dropped... Reconnecting...' + '🔄 Reconnected —
resuming…') with no provider name, so multi-provider sessions had to
dig through agent.log to attribute a drop. Replaced both call sites
with a single _emit_stream_drop helper that emits ONE line naming the
provider and error class, and always writes a structured WARNING to
agent.log with subagent_id, depth, provider, base_url, error_type.
Net UX change: 6 lines per triple-subagent drop → 3 lines, each
naming the provider. agent.log now has a structured breadcrumb per
retry that didn't exist before.
Tests: 6 new tests in tests/run_agent/test_stream_drop_logging.py
covering the logger-level guard, structured WARNING content, single
status line per drop (no Reconnected follow-up), and provider naming.
When the active main model has native vision and the provider supports
multimodal tool results (Anthropic, OpenAI Chat, Codex Responses, Gemini
3, OpenRouter, Nous), vision_analyze loads the image bytes and returns
them to the model as a multimodal tool-result envelope. The model then
sees the pixels directly on its next turn instead of receiving a lossy
text description from an auxiliary LLM.
Falls back to the legacy aux-LLM text path for non-vision models and
unverified providers.
Mirrors the architecture used in OpenCode, Claude Code, Codex CLI, and
Cline. All four converge on the same pattern: tool results carry image
content blocks for vision-capable provider/model combinations.
Changes
- tools/vision_tools.py: _vision_analyze_native fast path + provider
capability table (_supports_media_in_tool_results). Schema description
updated to reflect new behaviour.
- agent/codex_responses_adapter.py: function_call_output.output now
accepts the array form for multimodal tool results (was string-only).
Preflight validates input_text/input_image parts.
- agent/auxiliary_client.py: _RUNTIME_MAIN_PROVIDER/_MODEL globals so
tools see the live CLI/gateway override, not the stale config.yaml
default. set_runtime_main()/clear_runtime_main() helpers.
- run_agent.py: AIAgent.run_conversation calls set_runtime_main at turn
start so vision_analyze's fast-path check sees the actual runtime.
- tests/conftest.py: clear runtime-main override between tests.
Tests
- tests/tools/test_vision_native_fast_path.py: provider capability
table, envelope shape, fast-path gating (vision-capable model uses
fast path; non-vision model falls through to aux).
- tests/run_agent/test_codex_multimodal_tool_result.py: list tool
content becomes function_call_output.output array; preflight
preserves arrays and drops unknown part types.
Live verified
- Opus 4.6 + Sonnet 4.6 on OpenRouter: model calls vision_analyze on a
typed filepath, gets pixels back, reads exact text from images that
no aux description could capture (font color irony, multi-line
fruit-count list, etc.).
PR replaces the closed prior efforts (#16506 shipped the inbound user-
attached path; this PR closes the gap for tool-discovered images).
Adds test_notifier_second_blocked_delivers to cover the case where a
task is blocked, unblocked, then blocked again — the second blocked
event must still deliver a gateway notification.
Currently fails because blocked is treated as a terminal event kind,
causing the subscription to be dropped after the first block.
Linux's MAX_ARG_STRLEN caps any single argv element at 128 KB
(32 * PAGE_SIZE). The previous heredoc-in-the-command-string approach
in _write_to_sandbox put the entire tool result inside the 'bash -c'
arg, so any result over ~128 KB raised OSError [Errno 7] 'Argument
list too long' before the heredoc ever ran. The caller logged a
warning, but quiet_mode (CLI default) sets tools.* to ERROR — so the
warning never reached agent.log either, and the agent saw a 1.5 KB
preview tagged 'Full output could not be saved to sandbox'. Hits
delegate_task with 3+ subagent outputs routinely now.
Switch to passing content via env.execute(stdin_data=...). cmd is
now just 'mkdir -p X && cat > Y' (under 1 KB), and the heavyweight
payload travels through stdin where there is no argv-element limit.
E2E reproduced the user's exact 144,778-char delegate_task envelope:
old code OSError'd, new code round-trips cleanly to disk with all
three task summaries intact.
* feat(curator): show rename map (where skills went) in user-visible summary
The full data has always been on disk in REPORT.md, but the user-visible
curator summary (gateway 💾 line, CLI session-start panel,
`hermes curator status`) was counts-only — "consolidated 4 into 2
umbrellas" with no names. Users only discovered renames when something
they expected was gone.
New `_build_rename_summary()` formats the rename map and appends it to
`final_summary`:
auto: 1 marked stale; llm: consolidated 2 into 1, pruned 1
archived 3 skill(s):
• docx-extraction → document-tools
• pdf-extraction → document-tools
• old-stale-thing — pruned (stale)
full report: hermes curator status
Empty on no-op ticks (no archives), so most ticks add zero log noise.
Cap of 10 entries keeps agent.log readable when a 50-skill
consolidation lands; the full list is always in REPORT.md.
`hermes curator status` indents continuation lines so the multi-line
summary reads as one logical field.
5 new tests in tests/agent/test_curator_classification.py covering
empty / consolidation / pruning / cap / mixed cases.
* feat(curator): show recent run summary once on `hermes update`
The rename map is now visible from where users actually look — the
update flow they explicitly run, instead of just the live gateway log
or transient CLI session-start panel.
Behavior:
- After `hermes update`, if the most recent curator run produced a
rename map (multi-line summary) that the user hasn't seen yet, print
it once with a 'last run Xh ago' header and a one-time-message
footer.
- Stamp `last_run_summary_shown_at = last_run_at` after printing so
subsequent `hermes update` invocations are silent until a newer
curator run lands.
- Silent on no-op runs (single-line summary like 'auto: no changes;
llm: no change'). Still stamps shown so we don't reconsider on
every update.
- Silent when the curator has never run (the existing first-run
notice handles that case).
Output:
ℹ Skill curator — last run 4h ago
auto: 1 marked stale; llm: consolidated 2 into 1, pruned 1
archived 3 skill(s):
• docx-extraction → document-tools
• pdf-extraction → document-tools
• old-stale-thing — pruned (stale)
full report: hermes curator status
(This message shows once per curator run. View anytime: hermes curator status)
State migration:
- `_default_state()` gains `last_run_summary_shown_at: None`. Existing
state files lack the field; `.get()` returns None; the comparison
treats any prior run as 'not yet shown' and prints once on next
update. Self-healing.
Wiring:
- Both `hermes update` paths in main.py call the new
`_print_curator_recent_run_notice()` right after the existing
first-run notice. Best-effort try/except so a state-load bug
never breaks the update flow.
6 tests in tests/hermes_cli/test_curator_recent_run_notice.py:
no-run / single-line / multi-line / show-once / new-run-resets /
time-formatter buckets.
Follow-up test fix for #22693 — the existing test for ps-failure +
pid-file fallback needed the /proc walk path stubbed too since /proc
is now consulted first.
Salvage of NousResearch/hermes-agent#7622.
Docker images often lack procps so `ps` is unavailable. Try reading
/proc/*/cmdline first (works in any Linux container) and fall back to
`ps -A eww` only when /proc is not present. PermissionError on
individual PIDs is silently skipped.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
run_gateway() calls refresh_systemd_unit_if_needed() on every invocation
so restart settings stay current after exit-code-75 respawns. The
user-scope unit path resolves under Path.home() (NOT sandboxed by
conftest, only HERMES_HOME is), and generate_systemd_unit() bakes the
current HERMES_HOME into the unit's Environment= line.
Result: any test that exercises run_gateway() end-to-end on a real
Linux dev box silently rewrites the developer's installed
~/.config/systemd/user/hermes-gateway.service with a polluted
HERMES_HOME pointing at /tmp/pytest-of-<user>/.../hermes_test. On the
next reboot, systemd loads that unit, the gateway starts looking at an
empty tmp dir, and Telegram/Discord/etc. all show as 'No messaging
platforms enabled' even though the user's real config is fine. Three
tests in tests/hermes_cli/test_gateway.py hit this path:
test_run_gateway_exits_cleanly_on_keyboard_interrupt,
test_run_gateway_exits_nonzero_when_start_gateway_reports_failure, and
test_run_gateway_root_guard_has_escape_hatch.
Two-layer fix:
1. _install_fake_gateway_run helper (covers all four run_gateway() call
sites in test_gateway.py and any future ones) now also stubs
supports_systemd_services and refresh_systemd_unit_if_needed.
2. refresh_systemd_unit_if_needed() itself sniffs the generated unit
body for /pytest-of- and /hermes_test markers and refuses to write
when present. Defense in depth so a future test that bypasses the
helper still can't corrupt the dev's gateway. Tests that legitimately
exercise the refresh flow (test_run_gateway_refreshes_outdated_unit_on_boot)
patch generate_systemd_unit to return synthetic content that doesn't
carry those markers, so they keep working.
Adds test_refresh_refuses_to_bake_pytest_tmpdir_into_real_user_unit as a
regression test for the source-side guard.
RuntimeError('claude CLI turn timed out') from a local OpenAI-compatible
shim was falling through to FailoverReason.unknown, surfacing as 'Empty
response from model' and burning 3 retry slots on the same failing
endpoint. _classify_by_message had no timeout-message branch — only
billing/rate_limit/auth/context_overflow/model_not_found patterns. The
type-based check at line 565 also requires isinstance(error, (TimeoutError,
ConnectionError, OSError)) — a plain RuntimeError doesn't match.
Add _TIMEOUT_MESSAGE_PATTERNS for 'timed out', 'deadline exceeded',
'request timed out', 'operation timed out', 'upstream timed out', 'turn
timed out'. _classify_by_message returns FailoverReason.timeout (retryable=True)
when any pattern matches.
Salvage of #22664's classifier portion. The original PR also bundled a
fallback self-selection guard which is now redundant (already on main
via #22780) plus DeepSeek thinking and session_search fixes that are
their own separate concerns.
Follow-up to #22780 — fixes the still-broken classification of
generic-typed provider-shim timeouts that #22780's dedup didn't cover.
Fallback chain entries with 'api_key_env: ENV_VAR_NAME' weren't being
resolved by either the init-time fallback path (line ~1660) or the
runtime _try_activate_fallback path (line ~8045). Only literal
'api_key' was honored; the snake_case 'api_key_env' alias documented
elsewhere in the config was silently dropped, so a 'provider: custom'
fallback with base_url + api_key_env worked as primary but failed as
fallback with 'no endpoint credentials found' / 401.
Adds 'or fb.get("api_key_env")' to the existing 'key_env' lookup in
both call sites, with empty-string-to-None coercion so unset env vars
don't poison the resolver.
Salvage of #22665's fallback portion. The original PR also bundled
gateway-degrade-on-no-adapters changes (those land via the carve-out
in #22853 which is the same code) and run_agent.py memory-nudge
counter hydration (issue #22357 territory, not mentioned in the
title). Drops both bundled pieces; keeps just the api_key_env fix.
Closes#5392.
When connected_count == 0 AND enabled_platform_count > 0, the gateway
treated 'all adapters returned None' identically to 'all adapters
failed to connect' — both as fatal startup errors. The 'returned None'
case happens when imports fail silently or when adapters are present
in config but their dependencies aren't installed (e.g. discord.py
missing). Cron jobs and other gateway-runtime work would unnecessarily
fail to start.
Split: only return False when startup_retryable_errors is non-empty
(real connection attempt failed). When the list is empty AND enabled
> 0, log a warning and continue running, matching the 'no platforms
enabled' cron path.
Salvage of #22642's gateway slice. Drops the bundled run_agent.py
memory-nudge counter hydration block (issue #22357 territory) which
wasn't mentioned in the PR description.
Closes#5196.
Problem: terminal.docker_env set in config.yaml was silently ignored.
Docker containers never received the user-specified env vars.
Root cause: docker_env was missing from all three config→env bridging
maps (cli.py env_mappings, gateway/run.py _terminal_env_map,
hermes_cli/config.py _config_to_env_sync) and from the terminal_tool
_get_env_config() reader. _create_environment() consumed the key from
container_config correctly, but it was always {} because TERMINAL_DOCKER_ENV
was never set.
Also extend the list-serialisation branches in cli.py and gateway/run.py
to handle dict values via json.dumps (lists already used json.dumps;
plain str() on a dict produces undecodable output).
Fix:
- cli.py: add "docker_env": "TERMINAL_DOCKER_ENV" to env_mappings;
serialise dict values with json.dumps alongside existing list path
- gateway/run.py: same additions to _terminal_env_map and serialisation
- hermes_cli/config.py: add "terminal.docker_env": "TERMINAL_DOCKER_ENV"
to _config_to_env_sync so `hermes config set terminal.docker_env …`
persists to .env correctly
- tools/terminal_tool.py: add docker_env key to _get_env_config() reading
TERMINAL_DOCKER_ENV via _parse_env_var with default "{}"
Tests: add test_docker_env_is_bridged_everywhere to
tests/tools/test_terminal_config_env_sync.py — stash-verified: fails on
origin/main, passes with fix.
Fixes#20537
After Popen succeeds with os.setsid (detached process group), 5 things
happen with no try/except: Thread construction, reader.start(), lock
acquisition, prune+register, checkpoint write. If any raises, the
Popen object goes unregistered and the detached process group leaks
indefinitely.
Wrap the post-spawn setup in try/except. On failure:
- os.killpg(getpgid(pid), SIGKILL) takes down the entire process
group (not just the shell - important because of detached PG +
-lic shell wrapper that may have spawned children)
- proc.kill() fallback for ProcessLookupError/PermissionError/OSError
- proc.wait(timeout=5) reaps with a bound
- re-raise to preserve original traceback
Nested try/except around cleanup so a secondary failure can't mask the
original.
Closes#2749.
Problem
=======
`tools.checkpoint_manager._touch_project` reads the project metadata
file with `json.loads(meta_path.read_text(...))`, then immediately does:
meta["workdir"] = str(_normalize_path(working_dir))
The `except` block only catches `(OSError, ValueError)`. When the file
parses successfully but returns a non-dict value (a list `[]`, `null`,
or a scalar from a corrupted or hand-truncated write), `json.loads`
succeeds without error and `meta` is set to, e.g., `[]`. The subsequent
subscript assignment then raises `TypeError: list indices must be
integers or slices, not str`, which is NOT caught by the narrow except
clause.
This TypeError propagates up through `_take` to `ensure_checkpoint`,
where the broad `except Exception` safety net swallows it. The effect
is that `ensure_checkpoint` silently returns False for the entire
session — all checkpoints are skipped for the affected working directory
without any user-visible error.
Root cause
==========
Missing `isinstance(meta, dict)` guard after `json.loads`, identical in
pattern to bugs fixed in `cron/jobs.py` (#22569) and
`tools/process_registry.py` (#22544). The same guard is already
present one function below in `_list_projects` (line 506), but was
inadvertently omitted in `_touch_project`.
Fix
===
Add two lines after the try/except:
```python
if not isinstance(meta, dict):
meta = {}
```
This matches the existing guard in `_list_projects` and ensures a fresh
empty dict is used whenever the persisted value is not a mapping —
preserving the `created_at` semantics via `setdefault` on the next line.
Tests
=====
`TestTouchProjectMalformedMeta` covers four non-dict root values
(`[]`, `null`, `42`, `"oops"`). Each writes a corrupted metadata file,
calls `_touch_project`, and asserts: (a) no exception raised, (b) the
metadata file is rewritten as a valid dict containing `last_touch` and
`workdir`. All four fail on main with `TypeError`, pass with fix.
Full `tests/tools/test_checkpoint_manager.py` regression: 77 passed.
The FTS5 trigram tokenizer requires >=3 CJK characters per individual
token to produce matchable trigrams. A query like "广西 OR 桂林 OR 漓江"
has cjk_count=6 (passes the existing >=3 guard) but each token is only
2 CJK chars, so the trigram index returns 0 results.
Fix:
- Add per-token check: if any non-operator CJK token has <3 CJK chars,
force the LIKE fallback path regardless of total cjk_count.
- Expand the LIKE fallback to build one LIKE condition per non-operator
token joined with OR, so each term is matched independently.
Regression tests added in TestCJKSearchFallback:
- test_cjk_or_combined_short_tokens_returns_results
- test_cjk_short_token_or_query_preserves_filters
Problem:
When a provider or proxy drops a streaming response mid-flight (httpcore
raises RemoteProtocolError: "incomplete chunked read", "peer closed
connection", "response ended prematurely", etc.), _generate_summary
would not classify it as a transient error. Instead of retrying on the
main model, it entered the generic 60-second cooldown, leaving context
growing unbounded until the cooldown expired. Issue #18458.
Root cause:
_is_connection_error in auxiliary_client.py did not match httpcore's
streaming premature-close error substrings. context_compressor.py's
_generate_summary except block never called _is_connection_error, so
those errors fell through to the 60-second generic cooldown rather than
triggering the retry-on-main fallback path used for timeouts.
Fix:
1. auxiliary_client.py — extend _is_connection_error keyword list with:
"incomplete chunked read", "peer closed connection",
"response ended prematurely", "unexpected eof",
"remoteprotocolerror", "localprotocolerror".
Also guard the `from openai import ...` with try/except ImportError
so the function works in environments without the openai package.
2. context_compressor.py — import _is_connection_error and call it in
_generate_summary's except block as _is_streaming_closed. Include
_is_streaming_closed in the fallback-to-main condition (alongside
_is_model_not_found, _is_timeout, _is_json_decode) and use the
shorter 30s transient cooldown for streaming-closed errors.
Tests:
4 new regression tests in TestStreamingClosedFallback:
- test_incomplete_chunked_read_falls_back_to_main
- test_peer_closed_connection_falls_back_to_main
- test_streaming_closed_on_main_uses_short_cooldown (stash-verified)
- test_non_streaming_unknown_error_still_uses_long_cooldown
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
PR #2974 whitelisted three reasoning fields (reasoning, reasoning_details,
codex_reasoning_items) for the gateway's simple-text replay branch. Three
more fields were added to the DB later but the whitelist was never updated:
- reasoning_content: provider-facing thinking text. _copy_reasoning_content_for_api
promotes 'reasoning' -> 'reasoning_content' at send time only when the
strings happen to match. Carrying the original verbatim avoids loss
for providers that return them as distinct fields (DeepSeek/Kimi/
Moonshot thinking modes), and preserves the empty-string sentinel
that DeepSeek V4 Pro requires for thinking-mode replay.
- codex_message_items: exact assistant message items with 'phase'.
OpenAI docs: 'preserve and resend phase on all assistant messages —
dropping it can degrade performance.' Required for prefix cache hits.
No recovery path exists — once dropped, gone.
- finish_reason: informational; cheap to keep so transcripts replay
identically across CLI and gateway.
The CLI is unaffected because cli.py keeps the live in-memory message list
across turns (cli.py:10046 'self.conversation_history = result["messages"]').
The gateway rebuilds agent_history from the SQLite transcript on every turn,
so any field stripped during replay is silently lost.
Refactors the inline whitelist into a module-level _build_replay_entry()
helper so the contract can be unit-tested. 16 new tests pin the field set
and falsy-value handling.
Verified end-to-end: DB stores all 8 fields, replay now preserves all 8
(was preserving only 5 for assistant text turns).
Pick openrouter/pareto-code as your model and OpenRouter auto-routes each
request to the cheapest model meeting your coding-quality bar (ranked by
Artificial Analysis). The new openrouter.min_coding_score config key (0.0-1.0,
default 0.65) tunes the floor.
- hermes_cli/models.py: add openrouter/pareto-code to OPENROUTER_MODELS so
it shows up in the picker with a description
- hermes_cli/config.py: add openrouter.min_coding_score (default 0.65 — lands
on a mid-tier coder on the current Pareto frontier)
- plugins/model-providers/openrouter: emit extra_body.plugins =
[{id: pareto-router, min_coding_score: X}] when model is openrouter/pareto-code
AND the score is a valid float in [0.0, 1.0]
- agent/transports/chat_completions.py: same emission on the legacy flag
path (when no provider profile is loaded)
- run_agent.py: openrouter_min_coding_score kwarg + storage; plumbed into
both build_kwargs() invocations and the context-summary extra_body path
- cli.py: read openrouter.min_coding_score once at init, validate float in
[0,1], pass to AIAgent constructions (CLI + background-task paths)
- cron/scheduler.py, batch_runner.py, tools/delegate_tool.py,
tui_gateway/server.py: propagate the kwarg (mirrors providers_order
plumbing — subagents inherit, cron/batch read from config)
- tests: profile-level + transport-level coverage of the model gating,
unset/empty/out-of-range handling, and the legacy flag path
- docs: new 'OpenRouter Pareto Code Router' section in providers.md
Verified end-to-end against api.openrouter.ai: at score=0.65 we land on a
mid-tier coder, at omission we get the strongest. Score is silently dropped
on any model other than openrouter/pareto-code, so it's safe to leave set.
Pass session_id through to provider profile build_api_kwargs_extras so
the OpenRouter profile can attach an xAI cache-affinity header
(x-grok-conv-id: <session-id>) for x-ai/grok-* models. xAI prompt
cache requires server affinity via this header — without it the cache
is poisoned and Grok prompt-cache hit rates drop dramatically on
multi-turn sessions.
Carve-out of #22708 by Ninso112. The original PR bundled a /diff
slash command, a zsh completion fix (already on main via #22802),
and holographic memory null-guards. This salvage keeps just the
Grok header work — small, targeted, and well-tested. Other
contributors and changes preserved for separate review.
Closes#22705.
Add a configurable notifications mode for the Telegram platform adapter
that controls which messages trigger push notifications.
- display.platforms.telegram.notifications: "all" (default) | "important"
- HERMES_TELEGRAM_NOTIFICATIONS env var override
- In "important" mode, all sends use disable_notification=True except:
- Approvals (send_exec_approval) and slash confirmations
- Final response messages (metadata["notify"]=True)
- Zero overhead in default "all" mode
- Zero impact on non-Telegram platforms
Closes#22771
acp_command / acp_args descriptions previously primed the model to
populate them — "Per-task ACP command override (e.g. 'copilot')" —
even when no ACP CLI was installed. Models with weaker schema-following
discipline would set them and the spawn would fail.
Add explicit "Do NOT set unless the user has explicitly told you"
guidance at both the top-level acp_command and the per-task override.
Strengthen acp_args to mention it's empty unless acp_command is set.
Adds 2 tests pinning the descriptions.
Note: this is a cosmetic prompt-engineering fix — the params remain
exposed in the schema. The fully-correct fix is to gate them behind
a config flag or runtime ACP-CLI detection so the schema only emits
them when an ACP harness is available. Tracked as a follow-up; this
PR ships the low-cost stopgap.
Salvage of #22680 (delegate schema only). The original PR also
bundled unrelated fixes for #22548, #21944, #22150 — those
need separate PRs since #22548 and #21944 are already addressed
on main (#22780 + #22798 in flight) and #22150 deserves its own
review.
Closes#22013.
Two co-located fixes:
1. agent/model_metadata.py: bump hy3-preview static fallback from
256000 to 262144 (256 * 1024) to match OpenRouter live metadata
so cache and offline both agree (issue #22268).
2. tests/hermes_cli/test_tencent_tokenhub_provider.py: replace the
exact-value change-detector (assert ctx == 256000) with an
invariant assertion (registered + >= 4096). Per AGENTS.md
'Don't write change-detector tests': pinning the upstream-controlled
context length is exactly the test class the rule forbids — it
breaks every time the provider bumps the published value, with
zero behavioral coverage gained.
Salvage of #22574 with a redirect on the test approach. The
contributor's diff bumped the integer and added a SECOND
change-detector pinning DEFAULT_CONTEXT_LENGTHS[hy3-preview] == 262144,
which would re-break on the next published bump. We instead delete
the change-detector entirely and assert the relationship.
Closes#22268.
Problem
-------
`hermes doctor` ran two health checks for Anthropic: a dedicated one
with the correct `x-api-key` + `anthropic-version` headers, and a
generic Bearer-auth one driven by the pluggable `ProviderProfile` for
"anthropic". The generic check called `https://api.anthropic.com/v1/models`
with `Authorization: Bearer ...`, which Anthropic answers with HTTP 404,
producing a noisy duplicate warning even when the dedicated check passed.
Root cause
----------
`hermes_cli/doctor.py:_build_apikey_providers_list` deduplicated profiles
against a `_known_canonical` set built from the static list (Z.AI/GLM,
Kimi, DeepSeek, …). Providers with their own dedicated check above the
generic loop (Anthropic, OpenRouter, Bedrock) were not in that set, so
their profiles were appended and ran a second, broken check.
Fix
---
Add `{"anthropic", "openrouter", "bedrock"}` to the skip set, and
also skip profiles whose aliases match any of those names (e.g.
`claude`, `claude-oauth` → anthropic).
Tests
-----
tests/hermes_cli/test_doctor_dedicated_provider_skip.py:
- test_build_apikey_providers_list_skips_dedicated_check_providers:
asserts the assembled list does not contain anthropic, openrouter,
or bedrock entries.
- test_build_apikey_providers_list_includes_non_dedicated_providers:
sanity guard that legitimate providers (DeepSeek, Z.AI/GLM) survive.
Both confirmed via stash-verify (fail pre-fix with anthropic/openrouter
leaking, pass post-fix).
Fixes#22346
ALTER TABLE calls inside _migrate_add_optional_columns were guarded by a
snapshot of PRAGMA table_info taken at function entry. When the gateway
dispatcher opens the kanban DB twice per tick (once in _tick_once_for_board
and once via init_db's discard-and-reconnect path), a second connection can
run the same migration before the first one commits, causing:
sqlite3.OperationalError: duplicate column name: consecutive_failures
This crashed the dispatcher on every first tick after a gateway restart
(subsequent ticks succeeded because the columns were then present).
Fix: introduce _add_column_if_missing() which wraps ALTER TABLE in a
try/except that swallows OperationalError whose message contains
'duplicate column name'. All ALTER TABLE calls in
_migrate_add_optional_columns are routed through this helper.
Closes#21708
DeepSeek V4 Pro returns thinking content as typed blocks inside the
content array rather than as a top-level reasoning_content field:
[{"type": "thinking", "thinking": "..."}, {"type": "output", ...}]
_extract_reasoning only handled content as a plain string, so the
thinking text was silently dropped. On the next turn the session was
replayed without the thinking block, causing:
HTTP 400: The content[].thinking in the thinking mode must be
passed back to the API.
Fix: when content is a list and no structured reasoning field was
found, scan for items with type=='thinking' and accumulate their
'thinking' (or 'text') value into reasoning_parts. Structured fields
(reasoning, reasoning_content, reasoning_details) still take priority
so existing provider behaviour is unchanged.
Closes#21944
skills/media/youtube-content/scripts/fetch_transcript.py and
optional-skills/productivity/memento-flashcards/scripts/youtube_quiz.py
both import youtube-transcript-api at runtime, but the package was not
listed in pyproject.toml. A fresh `uv sync` therefore omits it, and
both skills fail on first invocation with:
ModuleNotFoundError: No module named 'youtube_transcript_api'
Add a new [youtube] optional-dependency group with
youtube-transcript-api>=1.2.0 (the v1.x API surface the scripts already
use) and include it in [all] so standard installs pick it up.
Regression tests: TestPyprojectDeclaresYoutubeExtra verifies the extra
is present in pyproject.toml and included in [all].
Closes#22243
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
163/NetEase IMAP servers reject every UID SEARCH/FETCH with `BYE Unsafe
Login` unless the client first identifies itself via the RFC 2971 ID
command after LOGIN. Without this, the email gateway logs in OK but
then fails on the very first poll and the connection is torn down.
Send the ID payload best-effort after both `imap.login()` sites
(`EmailAdapter.connect` and `_fetch_new_messages`). Failures are
swallowed at debug level so non-supporting IMAP servers (Gmail,
Outlook, Fastmail, Yahoo, etc.) keep working unchanged.
Closes#22271
Problem: `_get_cloud_provider()` set `_cloud_provider_resolved = True`
before resolution. If credentials were briefly unavailable on the first
call (e.g. a managed Nous Portal token mid-refresh), the resolver pinned
the entire process to local mode forever, even after credentials
self-healed seconds later.
Root cause: bookkeeping was set up-front, so any code path that fell
through to `return _cached_cloud_provider` (config read failure, no
credentials yet, explicit-provider instantiation failure) committed the
transient `None` to the cache permanently.
Fix: invert the bookkeeping. `_cloud_provider_resolved = True` is now
set only when (a) the user explicitly chose `cloud_provider: local`, or
(b) a provider was successfully resolved. All transient `None` paths
return without poisoning the cache, so the next call retries. Explicit
provider instantiation failures now log at warning level with stack
trace so operators can diagnose them.
Tests: 5 new cases in tests/tools/test_browser_cloud_provider_cache.py
covering explicit local, successful resolution, no-credentials-yet,
config read failure, and explicit provider instantiation failure.
Stash-verify confirmed the 3 transient-None tests fail without the fix.
All 320 existing browser tests still green.
Closes#22324
`fetch_models_dev()` is on the hot path of every `AIAgent.__init__`
(via `context_compressor → get_model_context_length`). The previous
policy was "always try network first, only fall back to disk if
network fails," so every fresh `hermes chat` / `hermes gateway` /
batch / cron process paid 250-500 ms re-fetching a 2 MB JSON registry
that was already on disk from earlier runs.
Add a stage 2 between in-mem and network: if
`models_dev_cache.json` exists and its mtime is younger than the
existing `_MODELS_DEV_CACHE_TTL` (1 hour, same TTL the in-mem cache
already uses), load from disk and skip the network call.
The in-mem TTL is anchored to the disk file's age, so a 50-min-old
cache stays in-memory for only 10 more minutes — no surprise
extension of staleness window.
Invariants preserved:
- `force_refresh=True` still always hits the network and only falls
back to disk on failure (`hermes config refresh` semantics).
- Missing disk cache → fall through to network (first-ever run).
- Stale disk cache (mtime > TTL) → fall through to network.
- Negative file age (clock skew) → fall through to network.
- Network failure → existing stage-4 stale-disk fallback unchanged.
Measured impact (3-run medians, 9950X3D, fresh process per run):
fetch_models_dev cold: 256 → 17 ms (-93%)
hermes chat -q wall: 4.00 → 3.73 s (-7% median)
3.99 → 3.60 s (-10% min)
The chat-end-to-end win is bounded below by API latency variance, but
the fetch_models_dev microbenchmark is the cleanest signal: 239 ms
shaved off every fresh-process agent construction.
Win compounds with the previous perf PRs:
#22681 google_chat lazy-load
#22766 doctor parallel + IMDS off
#22790 gateway.platforms PEP 562
Tests: all 30 `tests/agent/test_models_dev.py` pass (added 4 new ones
covering the new disk-cache-first path, force_refresh override, stale
disk fallback, and missing-disk-cache fall-through). Full `tests/agent/`
suite: 2560 passed, 0 failed.
The is_xai_responses branch only sent include=[reasoning.encrypted_content]
without forwarding the resolved reasoning_effort. Other Responses providers
(OpenAI, GitHub) already get effort forwarded — this aligns the xAI path.
Without this, agent.reasoning_effort is silently dropped on the xAI direct
path, making Hermes unable to control reasoning depth on grok-4.x via
api.x.ai. Tests added to TestCodexBuildKwargs cover effort passthrough,
disabled state, and minimal-clamp parity with non-xAI.
The xAI image-gen provider was DOA from PR #14765 onward — every request
422'd because the resolution param was being mapped to '1024'/'2048' but
xAI's API expects the literal strings '1k'/'2k'. PR #18678 fixed the
mapping; this test asserts the wire payload carries the literal so the
regression cannot recur silently.
Returning users who enabled '🖱️ Computer Use (macOS)' via 'hermes tools'
saw '✓ Saved configuration' but no install — cua-driver was never on
PATH and the toolset failed at first use. Two compounding causes:
1. _toolset_needs_configuration_prompt fell through to _toolset_has_keys,
which returned True for any provider with empty env_vars. cua-driver
has no env vars, so the gate skipped _configure_toolset entirely and
_run_post_setup('cua_driver') never ran.
2. No stable CLI entry-point existed for re-running the install when
the picker no-op'd it (e.g. when toggling the toolset off+on inside
one picker session, where 'added' is empty).
Changes:
- hermes_cli/tools_config.py: add _POST_SETUP_INSTALLED registry
mapping post_setup keys to installed-state predicates. The gate
now returns True when any visible provider has a registered
post_setup whose predicate fails. cua_driver is the only opt-in
for now; other post_setup hooks keep their existing behaviour.
- hermes_cli/main.py: add 'hermes computer-use install' and
'hermes computer-use status' as a stable docs target. install
reuses the same _run_post_setup('cua_driver') path that the
picker invokes; status reports whether cua-driver is on PATH.
- tools/computer_use/cua_backend.py: install hint now points users
at 'hermes computer-use install' first.
- website/docs/user-guide/features/computer-use.md: document the
new command as the primary install path.
- website/docs/reference/cli-commands.md: catalog 'hermes
computer-use' alongside 'hermes tools'.
- tests/hermes_cli/test_post_setup_gating.py: regression coverage
for the gate predicate (missing -> setup forced, installed ->
setup skipped, broken predicate -> non-blocking, unregistered
keys -> behaviour unchanged).
Fixes#22737. Reported by @f-trycua.
_try_activate_fallback() walked the chain by index without comparing
the candidate entry against the currently-failing backend. So a
misconfigured chain that listed the same provider+model as the primary,
or two custom_providers entries pointing at the same shim URL, would
loop the same failure 3x for the same backend.
After the fix, advance() skips:
- entries where (provider, model) match the current agent's
- entries with a base_url + model matching the current backend
(catches two custom_providers names pointing at the same shim)
Recursing through self._try_activate_fallback() continues to the next
chain entry; if everything matches, returns False and the caller
moves on without retrying the same broken path.
3 regression tests covering same-provider-same-model skip, same-base_url-
same-model skip, and the all-self-matching-returns-False exhaustion path.
Closes#22548 (the Hermes-side portion). The 120s timeout itself in
the downstream claude-cli shim is a deployment concern documented in
that issue's wherewolf87 comment.
Native Windows, WSL, SSH sessions, and Windows Terminal all send
Ctrl+Enter as bare LF (c-j). Hermes was binding c-j as submit on
every POSIX platform, so Ctrl+Enter submitted instead of inserting
a newline on those terminals. Reported in #22379.
Add _preserve_ctrl_enter_newline() predicate that detects the
environments where Ctrl+Enter must produce a newline (sys.platform
== 'win32', SSH_CONNECTION/SSH_CLIENT/SSH_TTY env, WT_SESSION,
WSL_DISTRO_NAME, /proc/version 'microsoft' marker). Gate the
c-j-as-submit binding off in those environments and gate the
c-j-as-newline handler on. Local POSIX TTYs without those markers
(docker exec, plain ssh from a Mac) keep c-j as submit so plain
Enter still works on thin PTYs.
Add install_ctrl_enter_alias() in hermes_cli/pt_input_extras.py
mapping the three CSI-u / modifyOtherKeys variants of Ctrl+Enter
('\x1b[13;5u', '\x1b[27;5;13~', '\x1b[27;5;13u') to the
(Escape, ControlM) tuple Alt+Enter produces. This lets Kitty /
mintty / xterm-with-modifyOtherKeys users over SSH get a Ctrl+Enter
newline through the existing Alt+Enter handler.
9 new tests + extended existing test_lf_enter_binds_to_submit_handler_posix
to cover bare-local vs SSH branches.
Closes#22379.
Non-streaming /v1/chat/completions wrapped any AIAgent result \u2014 including
partial/failed runs \u2014 as a successful 200 with finish_reason='stop' and
the internal failure string substituted into message.content. API
clients had no way to distinguish 'agent answered: X' from
'agent crashed and the X you see is its error message'.
After the fix:
- completed: True \u2192 200 finish_reason='stop' (unchanged)
- partial + truncated text \u2192 200 finish_reason='length' + hermes extras
- partial + no text / failed \u2192 502 OpenAI error envelope (SDKs raise)
- other failures \u2192 200 finish_reason='error' + hermes extras
Adds X-Hermes-Completed / X-Hermes-Partial / X-Hermes-Error headers
plus a 'hermes' extras object on partial responses for clients that
want the full picture.
Closes#22496.
Gateway creates a fresh AIAgent per inbound message in several common
scenarios: cache miss, idle eviction (1h TTL), config-signature
mismatch, process restart. A freshly-built AIAgent has
_turns_since_memory=0 and _user_turn_count=0, so the
memory.nudge_interval trigger ('_turns_since_memory >=
_memory_nudge_interval') can never be reached when these reconstructions
happen on roughly the cadence of the interval. A user can chat for hours
on Telegram without ever seeing a self-improvement review fire.
Reconstruct the counters from conversation_history at the top of
run_conversation(), right after the existing _hydrate_todo_store call.
Idempotent guard ('if self._user_turn_count == 0') means a cached agent
that already accumulated counters keeps them; only freshly-built agents
hydrate. Modulo arithmetic preserves the original 1-in-N cadence rather
than firing a review immediately on resume.
7 regression tests pinning the contract (mid-cycle history, modulo wrap,
idempotency, zero-interval skip, role==user filtering, production-code
anchor).
Closes#22357.
Operator-controlled HERMES_PROFILE values were rendered as
'**${author}** (${ts}):' — markdown bold with no provenance prefix.
Worker comment bodies render directly underneath. A misleading
profile name like 'hermes-system' or 'operator' could be misread by
the next worker as a system directive above attacker-influenced
content (confused-deputy primitive gated on operator misconfig).
The LLM-controlled author-forgery surface was already closed in
#22435 (author removed from KANBAN_COMMENT_SCHEMA). This is
defense-in-depth: render with an explicit 'comment from worker
`<author>` at <ts>:' prefix so even 'hermes-system' resolves to
'comment from worker `hermes-system` at ...' — parseable as
worker-comment metadata, not a system directive. Strip backticks
from author so they can't break out of the fence.
Update test_build_worker_context_caps_comments to count by body
regex since the rendered author line now also starts with
'comment '.
Closes#22452.
When session_id rotates (e.g. /new), commit_memory_session was firing
MemoryManager.on_session_end but skipping ContextEngine.on_session_end.
Engines that accumulate per-session state (LCM-style DAGs, summary
stores) leaked that state from the rotated-out session into whatever
continued under the same compressor instance.
Mirror the call shutdown_memory_provider already makes — same
lifecycle moment, same hook contract ("real session boundaries (CLI
exit, /reset, gateway expiry)"). /new is a real boundary for the old
session_id; providers keep their state but the rotated-out session_id
is done.
6 regression tests covering both-hooks-fire, no-memory-manager,
no-context-engine, both failure-tolerant paths.
Closes#22394.
- Restore allowed_chats gate before thread_id check so ignored_threads
applies universally (even to guest mentions).
- Compute _message_mentions_bot once in _should_process_message to
eliminate redundant second entity scan when guest_mode=true and the
message does not mention the bot.
- Remove redundant _is_group_chat from _is_guest_mention (caller already
verified the message is a group chat).
- Update _telegram_allowed_chats docstring to note guest_mode exception.
- Add test coverage: bot_command entity, text_mention entity,
caption_entities, and ignored_threads + guest_mode interaction.
- Add nik1t7n to AUTHOR_MAP.
The original PR placed 'pwd = pytest.importorskip("pwd")' on line 4
but 'import pytest' on line 9 — NameError on module load. Same for
test_file_sync_back.py. Plus, the in-function 'pwd = pytest.importorskip'
calls in test_auto_detected_root_is_rejected confused Python's scope
analysis (later 'import pytest' made pytest local everywhere in the
function) and caused UnboundLocalError. Drop the now-redundant
in-function importorskip calls and rely on the module-level guard.
The github-pr-workflow skill wraps the URL in double-quotes
('curl -H ... "https://api.github.com/..."'), which the original
allowlist regex (\s+https://api...) did not match. Without this,
the bundled github-pr-workflow skill is still blocked at every
cron tick despite #22605's fix landing for the bare-URL form.
Make the leading quote optional and add a regression test pinning
both single- and double-quoted forms.
