- Re-export _ssrf_redirect_guard from __init__.py
- Fix _parse_json @staticmethod using self._log_tag
- Update test_detect_message_type to call as instance method
- Fix mock.patch path for httpx.AsyncClient in adapter submodule
- Remove @staticmethod from _detect_message_type, _convert_silk_to_wav,
_convert_raw_to_wav, _convert_ffmpeg_to_wav so they can use self._log_tag
- Replace all remaining hardcoded "QQBot" log args with self._log_tag
- Downgrade STT routine flow logs (download, convert, success) from info to debug
- Keep warning level for actual failures (STT failed, ffmpeg error, empty transcript)
Three closely-related fixes for shutdown / lifecycle hygiene.
1. _release_running_agent_state(session_key) helper
----------------------------------------------------
Per-running-agent state lived in three dicts that drifted out of sync
across cleanup sites:
self._running_agents — AIAgent per session_key
self._running_agents_ts — start timestamp per session_key
self._busy_ack_ts — last busy-ack timestamp per session_key
Inventory before this PR:
8 sites: del self._running_agents[key]
— only 1 (stale-eviction) cleaned all three
— 1 cleaned _running_agents + _running_agents_ts only
— 6 cleaned _running_agents only
Each missed entry was a (str, float) tuple per session per gateway
lifetime — small, persistent, accumulates across thousands of
sessions over months. Per-platform leaks compounded.
This change adds a single helper that pops all three dicts in
lockstep, and replaces every bare 'del self._running_agents[key]'
site with it. Per-session state that PERSISTS across turns
(_session_model_overrides, _voice_mode, _pending_approvals,
_update_prompt_pending) is intentionally NOT touched here — those
have their own lifecycles tied to user actions, not turn boundaries.
2. _running_agents_ts cleared in _stop_impl
----------------------------------------
Was being missed alongside _running_agents.clear(); now included.
3. SessionDB close() in _stop_impl
---------------------------------
The SQLite WAL write lock stayed held by the old gateway connection
until Python actually exited — causing 'database is locked' errors
when --replace launched a new gateway against the same file. We
now explicitly close both self._db and self.session_store._db
inside _stop_impl, with try/except so a flaky close on one doesn't
block the other.
Tests
-----
tests/gateway/test_session_state_cleanup.py — 10 cases covering:
* helper pops all three dicts atomically
* idempotent on missing/empty keys
* preserves other sessions
* tolerates older runners without _busy_ack_ts attribute
* thread-safe under concurrent release
* regression guard: scans gateway/run.py and fails if a future
contributor reintroduces 'del self._running_agents[...]'
outside docstrings
* SessionDB close called on both holders during shutdown
* shutdown tolerates missing session_store
* shutdown tolerates close() raising on one db (other still closes)
Broader gateway suite: 3108 passed (vs 3100 on baseline) — failure
delta is +8 net passes; the 10 remaining failures are pre-existing
cross-test pollution / missing optional deps (matrix needs olm,
signal/telegram approval flake, dingtalk Mock wiring), all reproduce
on stashed baseline.
Telegram's MarkdownV2 has no table syntax — pipes get backslash-escaped
and tables render as noisy unaligned text. format_message now detects
GFM-style pipe tables (header row + delimiter row + optional body) and
wraps them in ``` fences before the existing MarkdownV2 conversion runs.
Telegram renders fenced code blocks as monospace preformatted text with
columns intact.
Tables already inside an existing code block are left alone. Plain
prose with pipes, lone '---' horizontal rules, and non-table content
are unaffected.
Closes the recurring community request to stop having to ask the agent
to re-render tables as code blocks manually.
Cuts shard-3 local runtime in half by neutralizing real wall-clock
waits across three classes of slow test:
## 1. Retry backoff mocks
- tests/run_agent/conftest.py (NEW): autouse fixture mocks
jittered_backoff to 0.0 so the `while time.time() < sleep_end`
busy-loop exits immediately. No global time.sleep mock (would
break threading tests).
- test_anthropic_error_handling, test_413_compression,
test_run_agent_codex_responses, test_fallback_model: per-file
fixtures mock time.sleep / asyncio.sleep for retry / compression
paths.
- test_retaindb_plugin: cap the retaindb module's bound time.sleep
to 0.05s via a per-test shim (background writer-thread retries
sleep 2s after errors; tests don't care about exact duration).
Plus replace arbitrary time.sleep(N) waits with short polling
loops bounded by deadline.
## 2. Subprocess sleeps in production code
- test_update_gateway_restart: mock time.sleep. Production code
does time.sleep(3) after `systemctl restart` to verify the
service survived. Tests mock subprocess.run \u2014 nothing actually
restarts \u2014 so the wait is dead time.
## 3. Network / IMDS timeouts (biggest single win)
- tests/conftest.py: add AWS_EC2_METADATA_DISABLED=true plus
AWS_METADATA_SERVICE_TIMEOUT=1 and ATTEMPTS=1. boto3 falls back
to IMDS (169.254.169.254) when no AWS creds are set. Any test
hitting has_aws_credentials() / resolve_aws_auth_env_var() (e.g.
test_status, test_setup_copilot_acp, anything that touches
provider auto-detect) burned ~2-4s waiting for that to time out.
- test_exit_cleanup_interrupt: explicitly mock
resolve_runtime_provider which was doing real network auto-detect
(~4s). Tests don't care about provider resolution \u2014 the agent
is already mocked.
- test_timezone: collapse the 3-test "TZ env in subprocess" suite
into 2 tests by checking both injection AND no-leak in the same
subprocess spawn (was 3 \u00d7 3.2s, now 2 \u00d7 4s).
## Validation
| Test | Before | After |
|---|---|---|
| test_anthropic_error_handling (8 tests) | ~80s | ~15s |
| test_413_compression (14 tests) | ~18s | 2.3s |
| test_retaindb_plugin (67 tests) | ~13s | 1.3s |
| test_status_includes_tavily_key | 4.0s | 0.05s |
| test_setup_copilot_acp_skips_same_provider_pool_step | 8.0s | 0.26s |
| test_update_gateway_restart (5 tests) | ~18s total | ~0.35s total |
| test_exit_cleanup_interrupt (2 tests) | 8s | 1.5s |
| **Matrix shard 3 local** | **108s** | **50s** |
No behavioral contract changed \u2014 tests still verify retry happens,
service restart logic runs, etc.; they just don't burn real seconds
waiting for it.
Supersedes PR #11779 (those changes are included here).
SessionStore._entries grew unbounded. Every unique
(platform, chat_id, thread_id, user_id) tuple ever seen was kept in
RAM and rewritten to sessions.json on every message. A Discord bot
in 100 servers x 100 channels x ~100 rotating users accumulates on
the order of 10^5 entries after a few months; each sessions.json
write becomes an O(n) fsync. Nothing trimmed this — there was no
TTL, no cap, no eviction path.
Changes
-------
* SessionStore.prune_old_entries(max_age_days) — drops entries whose
updated_at is older than the cutoff. Preserves:
- suspended entries (user paused them via /stop for later resume)
- entries with an active background process attached
Pruning is functionally identical to a natural reset-policy expiry:
SQLite transcript stays, session_key -> session_id mapping dropped,
returning user gets a fresh session.
* GatewayConfig.session_store_max_age_days (default 90; 0 disables).
Serialized in to_dict/from_dict, coerced from bad types / negatives
to safe defaults. No migration needed — missing field -> 90 days.
* _session_expiry_watcher calls prune_old_entries once per hour
(first tick is immediate). Uses the existing watcher loop so no
new background task is created.
Why not more aggressive
-----------------------
90 days is long enough that legitimate long-idle users (seasonal,
vacation, etc.) aren't surprised — pruning just means they get a
fresh session on return, same outcome they'd get from any other
reset-policy trigger. Admins can lower it via config; 0 disables.
Tests
-----
tests/gateway/test_session_store_prune.py — 17 cases covering:
* entry age based on updated_at, not created_at
* max_age_days=0 disables; negative coerces to 0
* suspended + active-process entries are skipped
* _save fires iff something was removed
* disk JSON reflects post-prune state
* thread safety against concurrent readers
* config field roundtrips + graceful fallback on bad values
* watcher gate logic (first tick prunes, subsequent within 1h don't)
119 broader session/gateway tests remain green.
Follow-up on the native NVIDIA NIM provider salvage. The original PR wired
PROVIDER_REGISTRY + HERMES_OVERLAYS correctly but missed several touchpoints
required for full parity with other OpenAI-compatible providers (xai,
huggingface, deepseek, zai).
Gaps closed:
- hermes_cli/main.py:
- Add 'nvidia' to the _model_flow_api_key_provider dispatch tuple so
selecting 'NVIDIA NIM' in `hermes model` actually runs the api-key
provider flow (previously fell through silently).
- Add 'nvidia' to `hermes chat --provider` argparse choices so the
documented test command (`hermes chat --provider nvidia --model ...`)
parses successfully.
- hermes_cli/config.py: Register NVIDIA_API_KEY and NVIDIA_BASE_URL in
OPTIONAL_ENV_VARS so setup wizard can prompt for them and they're
auto-added to the subprocess env blocklist.
- hermes_cli/doctor.py: Add NVIDIA NIM row to `_apikey_providers` so
`hermes doctor` probes https://integrate.api.nvidia.com/v1/models.
- hermes_cli/dump.py: Add NVIDIA_API_KEY → 'nvidia' mapping for
`hermes dump` credential masking.
- tests/tools/test_local_env_blocklist.py: Extend registry_vars fixture
with NVIDIA_API_KEY to verify it's blocked from leaking into subprocesses.
- agent/model_metadata.py: Add 'nemotron' → 131072 context-length entry
so all Nemotron variants get 128K context via substring match (rather
than falling back to MINIMUM_CONTEXT_LENGTH).
- hermes_cli/models.py: Fix hallucinated model ID
'nvidia/nemotron-3-nano-8b-a4b' → 'nvidia/nemotron-3-nano-30b-a3b'
(verified against live integrate.api.nvidia.com/v1/models catalog).
Expand curated list from 5 to 9 agentic models mapping to OpenRouter
defaults per provider-guide convention: add qwen3.5-397b-a17b,
deepseek-v3.2, llama-3.3-nemotron-super-49b-v1.5, gpt-oss-120b.
- cli-config.yaml.example: Document 'nvidia' provider option.
- scripts/release.py: Map asurla@nvidia.com → anniesurla in AUTHOR_MAP
for CI attribution.
E2E verified: `hermes chat --provider nvidia ...` now reaches NVIDIA's
endpoint (returns 401 with bogus key instead of argparse error);
`hermes doctor` detects NVIDIA NIM when NVIDIA_API_KEY is set.
Adds NVIDIA NIM as a first-class provider: ProviderConfig in
auth.py, HermesOverlay in providers.py, curated models
(Nemotron plus other open source models hosted on
build.nvidia.com), URL mapping in model_metadata.py, aliases
(nim, nvidia-nim, build-nvidia, nemotron), and env var tests.
Docs updated: providers page, quickstart table, fallback
providers table, and README provider list.
#4b1567f4 (anthhub) added qrcode to the messaging extra for Weixin's
QR login. The same package is needed by:
* hermes_cli/dingtalk_auth.py — QR device-flow auth shipped in #11574
* gateway/platforms/feishu.py:3962 — Feishu QR login
These extras are independent of [messaging] (users can install
hermes-agent[dingtalk] or hermes-agent[feishu] without [messaging]),
so the dep needs to be declared on each.
Pin matches anthhub's choice (>=7.0,<8) for consistency. The all
extra inherits from all three, so it picks up qrcode transitively.
Adds parallel tests to tests/test_project_metadata.py — same shape
as test_messaging_extra_includes_qrcode_for_weixin_setup.
Refs #9431.
Byte-level reasoning models (xiaomi/mimo-v2-pro, kimi, glm) can emit lone
surrogates in reasoning output. The proactive sanitizer walked content/
name/tool_calls but not extra fields like reasoning or the nested
reasoning_details array. Surrogates in those fields survived the
proactive pass, crashed json.dumps() in the OpenAI SDK, and the recovery
block's _sanitize_messages_surrogates(messages) call also didn't check
those fields — so 'found' was False, no retry happened, and after 3
attempts the user saw:
API call failed after 3 retries. 'utf-8' codec can't encode characters
in position N-M: surrogates not allowed
Changes:
- _sanitize_messages_surrogates: walk any extra string fields (reasoning,
reasoning_content, etc.) and recurse into nested dict/list values
(reasoning_details). Mirrors _sanitize_messages_non_ascii coverage
added in PR #10537.
- _sanitize_structure_surrogates: new recursive walker, mirror of
_sanitize_structure_non_ascii but for surrogate recovery.
- UnicodeEncodeError recovery block: also sanitize api_messages,
api_kwargs, and prefill_messages (not just the canonical messages
list — the API-copy carries reasoning_content transformed from
reasoning and that's what the SDK actually serializes). Always
retry on detected surrogate errors, not only when we found
something to strip — gate on error type per PR #10537's pattern.
Tests: extended tests/cli/test_surrogate_sanitization.py with
coverage for reasoning, reasoning_content, reasoning_details (flat
and deeply nested), structure walker, and an integration case that
reproduces the exact api_messages shape that was crashing.
The 'Thinking Budget Exhausted' user-facing error message advised users to
'set model.max_tokens in config.yaml'. That config key is documented but
intentionally not wired through to the API call in CLI/gateway paths — we
omit max_tokens by default so the inference server uses its full output
budget (llama-server -1=infinity, vLLM max_model_len-prompt_len, etc.).
Users followed the suggestion, saw no change, and kept filing bugs (see
closed#4404, #10917, #6955 and PRs #5001/#6080/#6446/#6707/#7075/#8804/
#10924/#11173/#11268 — all reporting the same misdirection).
Replace the misleading suggestion with an actionable one: switch models
via /model. Lowering reasoning effort remains the primary remediation.
* fix(tests): make AIAgent constructor calls self-contained (no env leakage)
Tests in tests/run_agent/ were constructing AIAgent() without passing
both api_key and base_url, then relying on leaked state from other
tests in the same xdist worker (or process-level env vars) to keep
provider resolution happy. Under hermetic conftest + pytest-split,
that state is gone and the tests fail with 'No LLM provider configured'.
Fix: pass both api_key and base_url explicitly on 47 AIAgent()
construction sites across 13 files. AIAgent.__init__ with both set
takes the direct-construction path (line 960 in run_agent.py) and
skips the resolver entirely.
One call site (test_none_base_url_passed_as_none) left alone — that
test asserts behavior for base_url=None specifically.
This is a prerequisite for any future matrix-split or stricter
isolation work, and lands cleanly on its own.
Validation:
- tests/run_agent/ full: 760 passed, 0 failed (local)
- Previously relied on cross-test pollution; now self-contained
* fix(tests): update opencode-go model order assertion to match kimi-k2.5-first
commit 78a74bb promoted kimi-k2.5 to first position in model suggestion
lists but didn't update this test, which has been failing on main since.
Reorder expected list to match the new canonical order.
Move moonshotai/kimi-k2.5 to position #1 in every model picker list:
- OPENROUTER_MODELS (with 'recommended' tag)
- _PROVIDER_MODELS: nous, kimi-coding, opencode-zen, opencode-go, alibaba, huggingface
- _model_flow_kimi() Coding Plan model list in main.py
kimi-coding-cn and moonshot lists already had kimi-k2.5 first.
Live turn rendering used to show the streaming assistant text as one
blob with tool calls pooled in a separate section below, so the live
view drifted from the reload view (which threads tool rows inline via
toTranscriptMessages). Model now mirrors reload:
- turnStore gains streamSegments (completed assistant chunks, each
with any tool rows that landed between its predecessor and itself)
and streamPendingTools (tool rows waiting for the next chunk)
- turnController.flushStreamingSegment() seals the current bufRef into
a segment when a new tool.start fires; pending tools get attached to
that next chunk so order matches reload hydration
- recordMessageComplete returns finalMessages instead of one payload,
so appendMessage gets the same shape for live-ending turns as for
reloaded ones
- appLayout renders segments before the progress/streaming area, and
the streaming message + pending-tools fallback carry whatever tools
arrived after the last assistant chunk
- useVirtualHistory: track last-seen ScrollBox metrics in a ref inside
the post-layout effect and bump ver when sticky/top/vp change — the
subscribe-based rearm was sufficient for fresh clicks but not for the
"hydrated mid-commit, measured empty, then metrics settle" path where
nothing re-triggered the hook until the next unrelated keystroke
- useSessionLifecycle: resume scrollToBottom from queueMicrotask to
setTimeout(..., 0) so the fresh transcript has a full task turn to
commit + measure before we try to land at the newest content
useVirtualHistory set up its useSyncExternalStore subscription during
the first render, when scrollRef.current was still null (the ScrollBox
ref attaches during commit, after render). Its useCallback for
subscribe had a stable scrollRef identity as its only dep, so it never
re-subscribed once the ref actually attached — the hook stayed stuck
with vp=0, top=0, no scroll subscription. Small sessions fit entirely
in cold-start so you didn't notice; big /resume sessions got sliced to
the last 40 items with a huge topSpacer and the viewport sat on empty
space until some unrelated state change (e.g. a keystroke) re-rendered
and finally read a real vp.
- flip a hasScrollRef flag in useLayoutEffect once the ref attaches and
add it to the subscribe useCallback deps so useSyncExternalStore
rearms with a real subscription
- on resume, scrollToBottom() after history hydrates so the ScrollBox
lands at the newest messages instead of scrollTop=0 (stickyScroll
doesn't auto-engage on the initial empty→full dump)
- drop inline `import()` type annotation in useSessionLifecycle (import
`PanelSection` at the top like everything else)
- include `panel` and `session.resumeById` in the useMainApp useMemo
deps now that the event handler depends on them
- wrap the derived `selected` range in a useMemo so it has stable
identity and stops invalidating the TextInput `rendered` memo every
render
- prettier re-sorting of a couple of export/import lines
- hermes-ink: export `withInkSuspended()` + `useExternalProcess()` that
pause/resume Ink around an arbitrary external process (built on the
existing enterAlternateScreen/exitAlternateScreen plumbing)
- tui: `launchHermesCommand(args)` spawns the `hermes` binary with
inherited stdio, with `HERMES_BIN` override for non-standard launches
- tui: `/model` and `/setup` slash commands invoke the CLI wizards
in-place, then re-preflight `setup.status` and auto-start a session on
success — no more exit-and-relaunch to finish first-run setup
- setup panel now advertises those slashes instead of only pointing
users back at the shell
- tui_gateway: new `setup.status` RPC that reuses CLI's
`_has_any_provider_configured()`, so the TUI can ask the same question
the CLI bootstrap asks before launching a session
- useSessionLifecycle: preflight `setup.status` before both `newSession`
and `resumeById`, and render a clear "Setup Required" panel when no
provider is configured instead of booting a session that immediately
fails with `agent init failed`
- createGatewayEventHandler: drop duplicate startup resume logic in
favor of the preflighted `resumeById`, and special-case the
no-provider agent-init error as a last-mile fallback to the same
setup panel
- add regression tests for both paths
- tui_gateway: route approvals through gateway callback (HERMES_GATEWAY_SESSION/
HERMES_EXEC_ASK) so dangerous commands emit approval.request instead of
silently falling through the CLI input() path and auto-denying
- approval UX: dedicated PromptZone between transcript and composer, safer
defaults (sel=0, numeric quick-picks, no Esc=deny), activity trail line,
outcome footer under the cost row
- text input: Ctrl+A select-all, real forward Delete, Ctrl+W always consumed
(fixes Ctrl+Backspace at cursor 0 inserting literal w)
- hermes-ink selection: swap synchronous onRender() for throttled
scheduleRender() on drag, and only notify React subscribers on presence
change — no more per-cell paint/subscribe spam
- useConfigSync: silence config.get polling failures instead of surfacing
'error: timeout: config.get' in the transcript
- Use certifi CA bundle for aiohttp SSL in qr_login(), start(), and
send_weixin_direct() to fix SSL verification failures against
Tencent's iLink server on macOS (Homebrew OpenSSL lacks system certs)
- Fix QR code data: encode qrcode_img_content (full liteapp URL) instead
of raw hex token — WeChat needs the full URL to resolve the scan
- Render ASCII QR on refresh so the user can re-scan without restarting
- Improve error message on QR render failure to show the actual exception
Tested on macOS (Apple Silicon, Homebrew Python 3.13)
iLink context_token has a limited TTL. When no user message has arrived
for an extended period (e.g. overnight), cron-initiated pushes fail with
errcode -14 (session timeout).
Tested that iLink accepts sends without context_token as a degraded
fallback, so we now automatically strip the expired token and retry
once. This keeps scheduled push messages (weather, digests, etc.)
working reliably without requiring a user message to refresh the
session first.
Changes:
- _send_text_chunk() catches iLinkDeliveryError with session-expired
errcode (-14) and retries without context_token
- Stale tokens are cleared from ContextTokenStore on session expiry
- All 34 existing weixin tests pass
Previously a message like `<@&1490963422786093149> help` would spawn a
thread literally named `<@&1490963422786093149> help`, exposing raw
Discord mention markers in the thread list. Only user mentions
(`<@id>`) were being stripped upstream — role mentions (`<@&id>`) and
channel mentions (`<#id>`) leaked through.
Fix: strip all three mention patterns in `_auto_create_thread` before
building the thread name. Collapse runs of whitespace left by the
removal. If the entire content was mention-only, fall back to 'Hermes'
instead of an empty title.
Fixes#6336.
Tests: two new regression guards in test_discord_slash_commands.py
covering mixed-mention content and mention-only content.
Free-response channels already bypassed the @mention gate so users could
chat inline with the bot, but auto-threading still fired on every
message — spinning off a thread per message and defeating the
lightweight-chat purpose.
Fix: fold `is_free_channel` into `skip_thread` so threading is skipped
whenever the channel is in DISCORD_FREE_RESPONSE_CHANNELS (via env or
discord.free_response_channels in config.yaml).
Net change: one line in _handle_message + one regression test.
Partially addresses #9399. Authored by @Hypn0sis (salvaged from PR #9650;
the bundled 'smart' auto-thread mode from that PR was dropped in favor
of deterministic true/false semantics).
* fix(gateway): bound _agent_cache with LRU cap + idle TTL eviction
The per-session AIAgent cache was unbounded. Each cached AIAgent holds
LLM clients, tool schemas, memory providers, and a conversation buffer.
In a long-lived gateway serving many chats/threads, cached agents
accumulated indefinitely — entries were only evicted on /new, /model,
or session reset.
Changes:
- Cache is now an OrderedDict so we can pop least-recently-used entries.
- _enforce_agent_cache_cap() pops entries beyond _AGENT_CACHE_MAX_SIZE=64
when a new agent is inserted. LRU order is refreshed via move_to_end()
on cache hits.
- _sweep_idle_cached_agents() evicts entries whose AIAgent has been idle
longer than _AGENT_CACHE_IDLE_TTL_SECS=3600s. Runs from the existing
_session_expiry_watcher so no new background task is created.
- The expiry watcher now also pops the cache entry after calling
_cleanup_agent_resources on a flushed session — previously the agent
was shut down but its reference stayed in the cache dict.
- Evicted agents have _cleanup_agent_resources() called on a daemon
thread so the cache lock isn't held during slow teardown.
Both tuning constants live at module scope so tests can monkeypatch
them without touching class state.
Tests: 7 new cases in test_agent_cache.py covering LRU eviction,
move_to_end refresh, cleanup thread dispatch, idle TTL sweep,
defensive handling of agents without _last_activity_ts, and plain-dict
test fixture tolerance.
* tweak: bump _AGENT_CACHE_MAX_SIZE 64 -> 128
* fix(gateway): never evict mid-turn agents; live spillover tests
The prior commit could tear down an active agent if its session_key
happened to be LRU when the cap was exceeded. AIAgent.close() kills
process_registry entries for the task, tears down the terminal
sandbox, closes the OpenAI client (sets self.client = None), and
cascades .close() into any active child subagents — all fatal if
the agent is still processing a turn.
Changes:
- _enforce_agent_cache_cap and _sweep_idle_cached_agents now look at
GatewayRunner._running_agents and skip any entry whose AIAgent
instance is present (identity via id(), so MagicMock doesn't
confuse lookup in tests). _AGENT_PENDING_SENTINEL is treated
as 'not active' since no real agent exists yet.
- Eviction only considers the LRU-excess window (first size-cap
entries). If an excess slot is held by a mid-turn agent, we skip
it WITHOUT compensating by evicting a newer entry. A freshly
inserted session (zero cache history) shouldn't be punished to
protect a long-lived one that happens to be busy.
- Cache may therefore stay transiently over cap when load spikes;
a WARNING is logged so operators can see it, and the next insert
re-runs the check after some turns have finished.
New tests (TestAgentCacheActiveSafety + TestAgentCacheSpilloverLive):
- Active LRU entry is skipped; no newer entry compensated
- Mixed active/idle excess window: only idle slots go
- All-active cache: no eviction, WARNING logged, all clients intact
- _AGENT_PENDING_SENTINEL doesn't block other evictions
- Idle-TTL sweep skips active agents
- End-to-end: active agent's .client survives eviction attempt
- Live fill-to-cap with real AIAgents, then spillover
- Live: CAP=4 all active + 1 newcomer — cache grows to 5, no teardown
- Live: 8 threads racing 160 inserts into CAP=16 — settles at 16
- Live: evicted session's next turn gets a fresh agent that works
30 tests pass (13 pre-existing + 17 new). Related gateway suites
(model switch, session reset, proxy, etc.) all green.
* fix(gateway): cache eviction preserves per-task state for session resume
The prior commits called AIAgent.close() on cache-evicted agents, which
tears down process_registry entries, terminal sandbox, and browser
daemon for that task_id — permanently. Fine for session-expiry (session
ended), wrong for cache eviction (session may resume).
Real-world scenario: a user leaves a Telegram session open for 2+ hours,
idle TTL evicts the cached AIAgent, user returns and sends a message.
Conversation history is preserved via SessionStore, but their terminal
sandbox (cwd, env vars, bg shells) and browser state were destroyed.
Fix: split the two cleanup modes.
close() Full teardown — session ended. Kills bg procs,
tears down terminal sandbox + browser daemon,
closes LLM client. Used by session-expiry,
/new, /reset (unchanged).
release_clients() Soft cleanup — session may resume. Closes
LLM client only. Leaves process_registry,
terminal sandbox, browser daemon intact
for the resuming agent to inherit via
shared task_id.
Gateway cache eviction (_enforce_agent_cache_cap, _sweep_idle_cached_agents)
now dispatches _release_evicted_agent_soft on the daemon thread instead
of _cleanup_agent_resources. All session-expiry call sites of
_cleanup_agent_resources are unchanged.
Tests (TestAgentCacheIdleResume, 5 new cases):
- release_clients does NOT call process_registry.kill_all
- release_clients does NOT call cleanup_vm / cleanup_browser
- release_clients DOES close the LLM client (agent.client is None after)
- close() vs release_clients() — semantic contract pinned
- Idle-evicted session's rebuild with same session_id gets same task_id
Updated test_cap_triggers_cleanup_thread to assert the soft path fires
and the hard path does NOT.
35 tests pass in test_agent_cache.py; 67 related tests green.
