When a Windows user relaunches Hermes while an in-app update is still
running (the desktop vanished with no progress and looks crashed), the
fresh instance spawns its own dashboard backend. That backend re-locks
the venv shim, the updater's straggler cleanup (force_kill_other_hermes
-> taskkill /F /T /IM hermes.exe) kills it, the launch dies with the 45s
"backend didn't come up" timeout, and the user relaunches into the same
trap -- an infinite respawn/kill loop (#50238).
Root cause: no mutual exclusion between an applying update and a fresh
desktop spawning its own local backend.
Fix: the updater publishes a HERMES_HOME/.hermes-update-in-progress
marker (pid + start time) for the whole run via an RAII drop-guard that
removes it on every exit path (success, early return, panic). A
freshly-launched desktop checks the marker before spawning its local
backend and PARKS until the update finishes -- then brings the backend
up itself (it is the surviving instance; the updater's own relaunch hits
the single-instance lock and quits). A stale marker (dead pid or past a
20-minute ceiling) is pruned so a crashed updater can never strand
future launches. No rogue backend spawns mid-update, so
force_kill_other_hermes has nothing legitimate to kill.
Marker parse/staleness logic is extracted to update-marker.cjs and
unit-tested; the Rust guard has unit tests; the Rust-write <-> JS-read
contract is E2E-verified.
In Telegram streaming, the typing indicator persisted through the slow
final rich-text/MarkdownV2 finalize edit, so the '...typing' bubble
lingered for seconds after the last streamed token. Add a one-shot
on_before_finalize hook to GatewayStreamConsumer, fired once when the
stream transitions into its finalization path, and wire it on both
Telegram streaming call sites to call pause_typing_for_chat() before
the final edit. Cover hook ordering and once-only behavior in tests.
Fixes#49712
Root cause of #49145: the Windows ZIP-update path did rmtree(dst) then
copytree(src, dst). If the copy failed partway — common on that path,
which only runs because file I/O is already flaky on the machine — the
directory was left deleted with nothing copied back. ui-tui/ vanishing
is what broke 'hermes --tui' (WinError 267), but the bug hit every
top-level directory.
_atomic_replace_dir stages the new copy into a sibling temp dir and only
swaps it in on full success, restoring the original on failure. A failed
update now leaves the live tree untouched instead of half-deleted.
The Windows update path can leave tracked ui-tui/ files deleted in the
working tree (HEAD intact). The guard now self-heals: when ui-tui/ is
missing in a git checkout, run `git restore -- ui-tui` and continue,
falling back to the printed manual-recovery steps only when git can't
recover it (no checkout / restore failed).
Builds on konsisumer's missing-workspace guard.
Answers a recurring plugin-author question: how to read the active
profile and drive Hermes from inside a hook callback when ctx._cli_ref
is None (gateway, hermes chat -q, and kanban-spawned worker sessions).
- Adds a 'Act from inside a hook' section to the plugin guide covering
ctx.profile_name and ctx.dispatch_tool as the session-agnostic APIs,
with a kanban_task_blocked example, and notes there is no in-process
slash-command bridge for headless workers (shell out via the terminal
tool instead).
- Adds the three kanban lifecycle hooks to the hook reference table with
their process semantics.
- Pins the contract with a regression test: ctx.dispatch_tool invokes a
tool handler with _cli_ref=None (worker/hook context).
Requested by @Smithangshu on Discord.
Follow-up on salvaged #50347: the event surface table was missing the
billing.step_up.verification switch case, and the File map omitted
lib/perfPane.tsx.
After a worker crash + reclaim + respawn, the board could show a task in the
Ready lane while its task_run was 'running' and the new worker was actively
executing (#36910). The dispatcher could then treat live work as available and
double-assign.
Root cause: the three reclaim paths (detect_crashed_workers,
release_stale_claims heartbeat-stale backstop, enforce_max_runtime) each
snapshot a task's worker_pid/claim_lock, do liveness work, then reset
tasks.status back to 'ready' with only a 'WHERE status=running' guard. If the
task was reclaimed AND re-claimed by a NEW worker in between (new run, new
claim_lock, live pid), the stale UPDATE clobbered the live task: status flipped
to 'ready' while the fresh run stayed 'running'. claim_task is the only writer
that sets status='running', so nothing put it back — permanent desync.
Fix: gate each reset on the snapshot's claim_lock (and worker_pid where
available) so it only fires when the task is still owned by the worker the
reclaim was computed for. A stale reclaim now no-ops (rowcount 0) instead of
desyncing a re-claimed task. Genuine crashes (lock still matches) reclaim
exactly as before.
This is the same race class the in-gateway dispatch lock (single-writer ticks)
mitigates, closed at the row level so a single dispatcher's fast
reclaim->respawn across two ticks is also safe.
Closes#36910.
Per @egilewski's audit on this PR, the security fix is behaviorally
correct but lacks focused regression coverage for the two traversal
vectors it closes. Adding tests now so the path-traversal guard
cannot silently regress.
* test_restore_rejects_snapshot_id_traversal -- exercises the
snapshot_id input guard with seven hostile values (parent
traversal, single parent, bare '.', bare '..', forward slash,
backslash, empty string). Each must return False without touching
the filesystem.
* test_restore_rejects_manifest_rel_traversal -- exercises the
manifest rel guard by injecting '../../outside.txt' into a real
snapshot's manifest.json, seeding a source payload at the escaped
path, and asserting the destination outside HERMES_HOME does not
exist after restore. This is the higher-value test of the pair --
verified locally that it fails without the fix in
restore_quick_snapshot (the escape destination gets written) and
passes with the fix in place.
The 67 pre-existing tests in test_backup.py continue to pass.
The gateway dispatcher captured kanban.auto_decompose ONCE at boot, so a user
who flipped it to false to STOP auto-decompose had no way to make that take
effect short of restarting the gateway. Reported (#49638): auto-decompose
created and launched tasks the user never intended (while they were still
typing the task description), and 'even Hermes Agent couldn't disable this
feature' — because the live config edit was silently ignored.
Auto-decompose is a safety toggle; turning it off must halt fan-out on the
next tick. The dispatcher now re-reads the flag (and auto_decompose_per_tick)
from config every tick via the extracted _resolve_auto_decompose_settings(),
which fails SAFE (disabled) on a config read error so a transient failure can
never re-enable a feature the user turned off.
Closes#49638.
connect() wrapped its entire body in an unbounded blocking flock(LOCK_EX) on
every call (_cross_process_init_lock). A single process stalled inside the
critical section — or a stale lock held by a wedged worker — blocked every
other connect(), including the long-lived gateway dispatcher's next-tick
connect, forever. No timeout, no traceback, no recovery: the board silently
stopped being worked until a manual restart (issue #36644).
Two fixes:
1. Fast-path skip: once THIS process has initialized a path, the expensive
first-open work (header validation, integrity probe, schema + additive
migrations) is already cached in _INITIALIZED_PATHS. The steady-state
connect has nothing for the cross-process lock to protect, so it now opens
the connection (WAL + pragmas) under only the cheap in-process _INIT_LOCK
and never touches the file lock. This removes the lock from the dispatcher's
hot path entirely — a stalled external 'hermes kanban list' can no longer
block ticks.
2. Bounded acquire: even on first-init, _cross_process_init_lock now retries a
non-blocking acquire up to a 10s deadline, then logs a WARNING and proceeds
WITHOUT the cross-process lock. Safe because the in-process _INIT_LOCK still
serializes same-process threads and the init work is idempotent
(CREATE TABLE IF NOT EXISTS + additive migrations) — worst case is redundant
work, not corruption. A bounded 'proceed anyway' beats an unbounded hang.
Windows path switched LK_LOCK -> LK_NBLCK (non-blocking) to match.
Closes#36644.
_default_spawn launched the worker subprocess with cwd=workspace and set
HERMES_KANBAN_WORKSPACE, but never set TERMINAL_CWD — so the worker inherited
the dispatching gateway's TERMINAL_CWD. That value takes precedence over the
process cwd in two places:
- tools/file_tools.py::_resolve_base_dir — a relative write_file path resolved
against the gateway user's home instead of the workspace, so artifacts
silently landed outside the workspace (#41312).
- agent_init's context-file loader — AGENTS.md was discovered relative to the
gateway's cwd, so under multi-profile dispatch a worker loaded whichever
gateway won the claim race's AGENTS.md, not the task's (#34619).
Both are the same root cause. Pinning TERMINAL_CWD to the workspace (where the
task's work actually happens) fixes both. Guarded on an existing absolute dir
because file_tools rejects relative/sentinel TERMINAL_CWD values — a non-dir
workspace leaves the inherited value rather than writing a meaningless one.
Closes#34619, closes#41312.
hermes -w created the worktree branch from the standalone clone's HEAD, which
lags origin when the clone isn't freshly updated (it's only refreshed by
hermes update, not per session). Every worktree branch then rooted on a stale
base, so the PR diff GitHub computes against current main ballooned with
unrelated changes and the agent had to discover the staleness at push time and
rebase.
_resolve_worktree_base() now fetches and branches from the freshest available
ref: the current branch's upstream if it tracks one (so a deliberate
feature-branch worktree tracks its own remote), else the remote's default
branch (origin/HEAD), else local HEAD as a fail-soft fallback (offline / no
remote / detached). A bogus 'origin/(unknown)' default is guarded, and worktree
creation retries from HEAD if branching off the remote ref fails — so this is
never worse than the old behavior.
Gated by worktree_sync (default true); set worktree_sync: false to keep the
old branch-from-local-HEAD behavior. The resolved base is printed in the
session banner.
This is the follow-up to the #50319 session, where the standalone clone was
213 commits behind origin and the worktree inherited that stale base.
Plugins could observe session/tool/approval lifecycle but had no way to
observe kanban task transitions. Adds three observer hooks fired by the
board's claim/complete/block transitions:
- kanban_task_claimed (dispatcher process, before worker spawn)
- kanban_task_completed (worker process, carries summary)
- kanban_task_blocked (worker process, carries reason)
Each fires AFTER the DB write txn commits, so a plugin observes durable
state and a slow/hanging callback can never hold the SQLite write lock.
All firing is best-effort: a raising hook is logged and swallowed and
never breaks a board transition. profile_name is resolved from
HERMES_HOME so dispatcher- and worker-side hooks carry the right profile.
Requested by @Smithangshu on Discord.
Plugins previously had no way to read the active profile name from the
PluginContext. The workaround in the wild — reaching into
ctx._manager._cli_ref — only works in an interactive CLI session;
_cli_ref is None in the gateway and in kanban-spawned worker sessions
(hermes -p <profile> chat -q ...), so the workaround breaks exactly
where multi-profile awareness matters most.
ctx.profile_name wraps hermes_cli.profiles.get_active_profile_name(),
which derives the name from HERMES_HOME and therefore works in every
execution context with zero dependency on _cli_ref.
After the agent's final response, the '...typing' bubble persisted ~5s.
send() re-triggers send_typing() after every delivery so the bubble
survives intermediate progress messages (Telegram clears typing on each
delivered message). But that re-trigger also fired on the FINAL send,
re-arming Telegram's ~5s timer AFTER the gateway had already torn down
its typing-refresh loop — and Telegram exposes no stop-typing API, so
nothing cancelled it.
Gate the post-send re-trigger on the absence of metadata['notify'] (set
only on the final user-visible reply via _mark_notify_metadata). Both
the rich-message and legacy send paths are covered; intermediate
progress sends still re-trigger so the bubble stays alive mid-response.
Fixes#48678
Add a platform-neutral send-failure vocabulary so consumers can branch on a
typed category instead of substring-matching the raw provider message.
- base.py: SEND_ERROR_KINDS + classify_send_error() (too_long / bad_format /
forbidden / not_found / rate_limited / transient / unknown), and an optional
SendResult.error_kind field (defaults None — fully backward compatible).
- telegram.py: populate error_kind on send() failures; message_too_long keeps
its existing error token plus error_kind='too_long'.
Purely additive: no behavioral change to the existing degrade-and-deliver
paths (MarkdownV2->plain-text fallback, overflow split, retry classification
all untouched). 22 new tests + 210 adapter regression tests green.
The port-announcement clock in waitForDashboardPort starts the instant the
backend process is spawned — before uvicorn binds its socket. On a cold
install the child first compiles and imports the whole hermes_cli.main ->
web_server -> FastAPI/uvicorn chain, and on Windows real-time AV scans every
freshly written .pyc. That pre-bind cost can exceed the old hardcoded 45s
deadline, so the desktop killed a healthy-but-still-starting backend and
respawned it, piling up orphaned processes (#50209).
Raise the default to 90s and make it overridable via
HERMES_DESKTOP_PORT_ANNOUNCE_TIMEOUT_MS, clamped to a 45s floor so a bad
override can't reintroduce the loop. Warm starts still announce in well under
a second; both call sites inherit the new default with no change. Adds
backend-ready.test.cjs (wired into test:desktop:platforms).
Three tests covering the scenarios from issue #50209 that could not be
validated with real Defender on a fresh install:
1. test_lifespan_warmup_is_nonblocking
Patches _warm_gateway_module to sleep 3 s. Measures TestClient startup
time — must complete in < 1.5 s, proving the fire-and-forget
run_in_executor does not block the event loop before port binding
(HERMES_DASHBOARD_READY timing proxy).
2. test_get_status_does_not_block_event_loop
Patches _resolve_restart_drain_timeout to sleep 3 s. Fires concurrent
GET /api/status and GET /api/version requests. /api/version must
respond in < 3 s while /api/status waits — proving the event loop
stays free during the slow import (15 s socket timeout would not fire).
3. test_concurrent_status_probes_all_respond
Three simultaneous /api/status probes with the slow patch — all must
return HTTP 200 (no connection resets, no orphan accumulation).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Fixes a regression introduced by the prior approach (synchronous import
hermes_cli.gateway inside _lifespan) that caused a new failure mode:
the blocking import stalled the asyncio event loop before uvicorn could
bind its port, pushing HERMES_DASHBOARD_READY past the desktop shell's
45 s announcement deadline and triggering a respawn loop that accumulated
orphaned backend processes.
Two-part fix:
_lifespan: replace the blocking import with a fire-and-forget
run_in_executor call (_warm_gateway_module). The import runs in a
worker thread while the server socket is already open, so
HERMES_DASHBOARD_READY fires without delay.
get_status: replace the inline lazy import with
await run_in_executor(None, _resolve_restart_drain_timeout). This is
the root fix for the original 15 s socket-timeout: the blocking
.pyc-compilation + Defender scan is offloaded to a thread, keeping the
event loop free for every /api/status probe. After the first call the
module is in sys.modules and the executor returns in microseconds.
Both helpers are extracted as module-level sync functions so they can
be unit-tested independently of FastAPI or uvicorn.
Closes#50209
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
On resource-contended hosts the embedded Hindsight daemon can exceed a
single 2s /health check; upstream then waits a grace window before
treating it as stale and killing+restarting it (hindsight-embed reads
HINDSIGHT_EMBED_PORT_HEALTH_GRACE_TIMEOUT, default 30s, into a
module-level constant at import time). Users on busy boxes had no
Hermes-side way to raise it short of hand-setting an env var.
Add a 'port_health_grace_timeout' config.json option to the Hindsight
plugin. When set, initialize() exports it to the process env BEFORE
daemon_embed_manager is imported (the import-time read is the contract).
setdefault() so an explicit operator env override always wins. Exposed
in 'hermes memory setup' for local_embedded mode.
Follow-up to #50308 / issue #13125 comment thread.
The read_file device guard now walks symlink hops before the file operation
layer, but that hop walk still interpreted relative paths against the Python
process cwd. In sessions where TERMINAL_CWD points at the task workspace, a
relative workspace symlink to a blocked alias such as /dev/../dev/stdin could
therefore miss the intermediate device target before later task-cwd resolution.
Anchor relative device checks to the task base before symlink-hop inspection so
the pre-I/O guard sees the same workspace path that read_file would otherwise
read. Absolute device paths and the existing final realpath fallback remain
unchanged.
Refs #10141
Refs #29158
Follow-up for salvaged PR #50256. Unit tests for the three behaviors:
retryable classification of Envoy/sidecar overflow strings, per-chat typing
cooldown with stop_typing reset, and the _supervise_sidecar crash-detection
path that raises a retryable fatal (and the clean-shutdown no-op).
When the Node spectrum-ts sidecar process exited mid-session (crash,
OOM, upstream overflow escalation), _supervise_sidecar returned
silently — readline hit EOF, the log-pump loop broke, and nothing
notified the gateway. _inbound_loop entered an infinite retry loop
against a dead port, _running stayed True, and the adapter remained
in self.adapters with no path to self-recovery short of a manual
gateway restart.
Add a death-detection tail to _supervise_sidecar: after the log-pump
exits (EOF or exception), guard on _inbound_running to distinguish
unexpected death from a deliberate disconnect(). On unexpected exit,
call _set_fatal_error("SIDECAR_CRASHED", retryable=True) followed by
_notify_fatal_error() so the reconnect watcher picks up the platform
within 30 s and retries with exponential backoff (30 s → 300 s cap)
until the sidecar comes back up. All other platforms remain unaffected.
The _inbound_running guard is safe against races: disconnect() sets
_inbound_running = False before _stop_sidecar() cancels the supervisor
task. CancelledError is BaseException, not Exception, so it bypasses
the except clause and propagates normally — the detection block never
runs during a clean shutdown.
Closes#50185
Two independent gaps let a transient Photon/Spectrum upstream overflow
degrade message delivery and amplify gRPC pressure:
1. _is_retryable_error did not recognise Photon- or Envoy-specific error
strings ("internal sidecar error", "upstream connect error",
"reset reason: overflow"), so _send_with_retry fell through to the
plain-text fallback immediately instead of backing off and retrying.
2. send_typing had no rate gate, so a burst of typing-indicator calls
during an overflow event kept hitting the upstream gRPC connection and
widened the failure window.
Fix:
- Add _PHOTON_RETRYABLE_PATTERNS with the three high-specificity Envoy /
sidecar substrings and override _is_retryable_error on PhotonAdapter to
check them after delegating to the base-class patterns. base.py and all
other adapters are untouched.
- Add a 5 s per-chat cooldown in send_typing backed by _typing_last_sent.
stop_typing clears the entry so the next start after a completed turn
fires immediately — only rapid consecutive starts without a stop are
suppressed.
- Reduce PhotonAdapter._send_with_retry default max_retries from 2 to 1
(single 2 s back-off check) — enough to confirm whether the Envoy
circuit-breaker has opened, without adding unnecessary latency.
All changes are scoped to plugins/platforms/photon/adapter.py.
* fix(api-server): stop silently promising async delivery on stateless HTTP path
terminal(notify_on_complete=True / watch_patterns) and delegate_task(background=True)
silently no-op'd on the API server / WebUI path (#10760): the watcher / detached
child registered, but every API-server route (OpenAI-spec /v1/chat/completions
and /v1/responses, plus the proprietary /v1/runs SSE stream) tears down its
channel when the turn ends, and APIServerAdapter.send() is a no-op stub. A
completion that fires after the response closed had nowhere to go — from the
agent side, indistinguishable from a hang.
There is no spec-compliant surface to wake the agent later on a stateless HTTP
client, so make the no-op honest instead of silent:
- Add a per-adapter capability flag supports_async_delivery (default True;
APIServerAdapter = False), propagated into a HERMES_SESSION_ASYNC_DELIVERY
contextvar via async_delivery_supported(). Toggle on the adapter, not a
hardcoded platform string — a future stateless adapter is correct-by-default.
- terminal: when delivery is unsupported, skip watcher registration, force
notify_on_complete off, and return a notify_unsupported note telling the
agent to process(action='poll').
- delegate_task: when delivery is unsupported, fall back to SYNCHRONOUS
execution (work runs and returns in the same response) with a note, instead
of handing out a handle that never resolves.
CLI (in-process completion_queue) and the real gateway platforms are unchanged.
Fixes#10760
* refactor(api-server): route session binding through a single no-delivery chokepoint
Add APIServerAdapter._bind_api_server_session() and route both agent-entry
paths (_run_agent for /v1/chat/completions + /v1/responses, and the /v1/runs
_run_sync path) through it. The helper hardwires platform="api_server" and
async_delivery=False with no async_delivery parameter to pass, so a future
route added to the API server physically cannot reintroduce the silent
no-op (#10760) by forgetting to mark the channel as non-delivering.
The binding stays request-scoped (cleared per turn), so a session resumed
later on a delivering interface (CLI / gateway platform) re-binds fresh and
is NOT blocked — the no-delivery decision tracks the interface handling the
current turn, never the session.
_collect_delegate_child_ids() walks the _delegate_from marker chain to
gather delegate subagents for cascade deletion, but started its visited
set empty. When the chain loops back onto a parent — a delegation cycle,
or a parent that is also another parent's delegate child when several ids
are deleted together — that parent was collected as one of its own
descendants and then permanently deleted, along with all of its messages,
by _delete_delegate_children().
Seed the visited set with the parent ids so they can never be re-collected,
and exclude them from the returned child set. Callers (delete_session,
bulk delete) remove the parents separately, so this only prevents the
unintended parent deletion; legitimate child collection is unchanged.
Add regression tests (in-memory sqlite) covering single/multi-level
delegate chains, the parent_session_id+marker branch, untagged children
(orphan-don't-delete contract), and the cycle case that previously leaked
the parent into the deletion set.
Fixes#49148
A shell-launched 'hermes gateway run --replace' / 'gateway restart' on a
systemd/launchd host can leave an orphan gateway whose kanban dispatcher
escapes the service cgroup, survives 'systemctl restart', and becomes a
second long-lived writer on the shared kanban.db. Two dispatchers that each
believe they own the file both pass SQLite busy_timeout and then race on WAL
frames — the documented root cause of multi-writer corruption (issue #35240).
The existing _guard_supervised_gateway_conflict startup guard blocks the
common way an orphan is born, but does nothing once a second dispatcher
already exists. This adds the defense-in-depth: dispatch_once now wraps every
tick in a non-blocking, board-scoped flock (_dispatch_tick_lock). A losing
dispatcher returns DispatchResult(skipped_locked=True) and does zero DB writes
this tick — so two dispatchers can never run a reclaim/spawn/write sequence
concurrently regardless of how the second one got there.
- Non-blocking (LOCK_NB): never stalls the gateway's async watcher.
- Board-scoped: lock file is a .dispatch.lock sibling of each board's
kanban.db, so unrelated boards tick in parallel.
- POSIX + Windows (fcntl / msvcrt LK_NBLCK), no-op degrade where neither
exists — mirrors the existing _cross_process_init_lock pattern.
Verified with a real two-process orphan repro: while a separate process holds
the lock, dispatch_once skips; after release it runs.
hermes backup only walks HERMES_HOME, so memory providers that keep
config/credentials in home-anchored dotdirs (honcho -> ~/.honcho,
hindsight -> ~/.hindsight, openviking -> ~/.openviking) lost that data
across a backup/import cycle — the peer IDs, session pairings, and API
keys never made it into the archive.
Add an optional MemoryProvider.backup_paths() hook (default []). The
active provider declares its external paths; backup resolves them from
config only (no init, no network), archives the ones under the home dir
into a reserved _external/ subtree encoded relative to home, and import
restores them to their original location with a home-anchored traversal
guard and 0600 on credential-shaped files. Paths outside home are
skipped as non-portable.
honcho, hindsight, and openviking override the hook. E2E-validated full
backup->import cycle plus 7 new tests.
Rich messages are not ready for primetime: current Telegram clients can
render Bot API 10.1 rich messages as blank/unsupported bubbles and make
them hard to copy as plain text, which is worse than the legacy
MarkdownV2 path for command snippets and mobile handoffs. Default the
rich_messages toggle to False so replies stay on the copyable legacy
path; users opt in per bot via platforms.telegram.extra.rich_messages:
true. Updates adapter, gateway config default, example config, English +
zh-Hans docs, and the default/opt-in tests.
A turn forcibly interrupted by the drain-timeout escalation never reaches
turn_finalizer.finalize_turn (the only place that flushes the turn to
state.db). Its in-flight tool rounds live only in the in-memory
_session_messages, so the immediate pre-restart turn was silently dropped
from load_transcript() on resume.
_finalize_shutdown_agents now flushes _session_messages to the SQLite
session store before teardown. The flush is idempotent (identity-tracked
in _flush_messages_to_session_db), so agents that finished gracefully
re-flush nothing. The resume_pending / fresh-tool-tail branches in
_handle_message_with_agent already expect a transcript whose tail may be a
pending tool result.
Fixes#13121.
Inbound image/audio/video payloads were buffered fully into process memory
before being written to the cache, with no size limit. A large upload
(Discord Nitro allows 500 MB) or a remote media URL in an inbound message
pointing at a huge file could spike RAM and OOM-kill the gateway.
Enforce a configurable cap in the shared cache helpers (gateway/platforms/
base.py) so the protection holds across every platform adapter, not one:
- cache_image/audio/video_from_bytes reject oversized payloads before writing
(video was the gap in the original report — now covered).
- cache_image/audio_from_url stream the body, rejecting on an oversized
Content-Length header and re-checking the running total per chunk so an
absent/lying header can't smuggle an unbounded body past the cap.
- Discord's _read_attachment_bytes checks att.size up front, so an oversized
attachment is rejected before any bytes are pulled into memory.
Configurable via gateway.max_inbound_media_bytes in config.yaml (default
128 MiB; 0 disables). No new env var — non-secret config lives in config.yaml.
Salvaged and extended from @sgaofen's PR #13341 (the original report and the
shared-helper approach). Reapplied onto current main (Discord adapter has
since moved to plugins/platforms/discord/), the configurable knob moved from
an env var to config.yaml, and the video cache helper added.
Co-authored-by: Hermes Agent <noreply@nousresearch.com>
The salvaged #19820 unifies the write_file guard under
_is_internal_file_tool_content with the message 'internal read_file
display text'. Two tests added to test_file_read_guards.py after the PR
branch point still asserted the old 'status text' wording. Update them
to match the new (correct, more general) message.
hermes config show printed the model dict raw via print(), bypassing the
logging redactor; a custom-provider api_key (e.g. Cloudflare cfut_...) was
shown in plaintext even with security.redact_secrets=true. Opaque tokens
don't match any vendor-prefix regex, so structural key-name masking is
required.
- Add redact_config_value(): recursively masks credential-shaped keys
(api_key/token/secret/... exact-match) via mask_secret.
- Wrap the show_config model dump in it.
- Mask the set_config_value echo when the leaf key is credential-shaped
(config set model.api_key routes to config.yaml, lowercase misses the
.env allowlist).
Bedrock Claude routes through the AnthropicBedrock SDK and injects
cache_control, so cached tokens are always reported — but the pricing
table had no cache cost fields for any Bedrock model, so /usage showed
"cost unknown" on every cached session. Also, cross-region inference
profiles (us./global./eu. prefixes) never matched the bare pricing keys.
- Add cache_read/cache_write rates to the four Bedrock Claude rows
(read 0.1x input, write 1.25x input per the Bedrock pricing page).
- Normalize the cross-region prefix in the Bedrock pricing lookup,
mirroring is_anthropic_bedrock_model's prefix list.
Closes#50295.
PostgreSQL's initdb refuses to run as root, so the embedded Hindsight
daemon could never initialize its data directory under root. The
daemon-start thread would fail, retry, and loop forever — each cycle
reloading embedding models (~958MB RAM, ~33% CPU) with no user-visible
error, leaving Hermes sluggish on a common VPS/cloud root setup.
initialize() now detects root (os.geteuid() == 0) before spawning the
daemon thread, disables local_embedded mode, and surfaces a clear
warning to both the log and the terminal so the user knows to run as a
non-root user or switch to cloud / local_external mode.
Closes#13125.
Co-authored-by: teknium1 <127238744+teknium1@users.noreply.github.com>
The TUI /compress slash side-effect compressed the session, synced the
key, and emitted session.info — but returned an empty string, so the
user saw no 'Compressed: N → M messages / ~X → ~Y tokens' feedback. The
CLI (_manual_compress) and gateway (slash_commands) paths both already
call summarize_manual_compression; the TUI slash path was the lone gap.
Snapshot history + rough token estimate before and after compaction and
return the formatted summarize_manual_compression() feedback, mirroring
the session.compress RPC handler. The estimate uses the same
estimate_request_tokens_rough(system_prompt, tools) inputs as the RPC
path, re-reading the system prompt after compaction (it may be rebuilt).
Co-authored-by: liuhao1024 <sunsky.lau@gmail.com>
The background-review fork (fires ~every 10 turns) pins
review_agent.session_id = agent.session_id — the parent's LIVE id — for
prefix-cache parity, then calls close(). With session finalization now in
close(), that would end the still-active parent session mid-conversation.
Set _end_session_on_close = False on the fork so the real owner (CLI close /
gateway reset / cron) finalizes the session instead.
Follow-up to the #12029 fix.
Funnel session finalization through AIAgent.close() — the single terminal
path every agent (CLI, gateway, subagent, cron) funnels through — so finished
agents stop leaving rows with ended_at IS NULL. The biggest leak source was
delegate_task subagent + background-review forks whose close() never ended
their row.
end_session() is first-reason-wins and no-ops on an already-ended row, so a
'compression'/'cron_complete'/'cli_close' reason set by an earlier terminal
path is never clobbered. /resume already calls reopen_session(), so
finalizing-on-close does not break resumability.
Temporary helper agents that rotate/share the session forward (manual
compression, gateway session-hygiene) opt out via _end_session_on_close=False.
Also stop the long-running gateway heartbeat once the executor is done or the
session slot is rebound to a different agent, preventing a stale
'running: delegate_task' bubble from outliving its run.
Closes#12029.
The 'Session compressed N times — accuracy may degrade' warning went
through _vprint (CLI stdout only), so the Ink TUI / Telegram / Discord
never saw it — unlike the two other compression warnings in the same
module, which route through _emit_status (and store _compression_warning
for late-bound gateway status_callback replay).
Set agent._compression_warning + call agent._emit_status() for this
warning too, matching the sibling pattern. _emit_status still _vprints
for the CLI, so CLI output is unchanged; TUI / gateway surfaces now
receive it via status_callback (and replay_compression_warning can
re-deliver it once a late-bound gateway callback is wired).
Co-authored-by: liuhao1024 <sunsky.lau@gmail.com>
When an OpenAI-compatible proxy (e.g. cmkey.cn, one-api Anthropic channels)
returns a well-formed HTTP 200 whose summary content is null or empty/
whitespace-only, _generate_summary coerced it to "" and stored a prefix-only
summary — silently replacing the compacted turns with nothing. The model then
lost all in-progress context after compression (#11978, #11914).
_validate_llm_response already guards None / empty-choices, so those never
reach the compressor; the gap was a well-formed response with empty *content*.
Now treat empty content as a summary failure: raise so it routes through the
existing main-model fallback then transient cooldown, dropping the turns
without a summary rather than wiping context with an empty one.
Also narrow the bare 'except RuntimeError' so only genuine 'No LLM provider
configured' errors take the 600s no-provider cooldown; empty/invalid-response
RuntimeErrors from a configured provider now correctly get the main-model
fallback instead of being misrouted into the long no-provider cooldown.
Reported by @Hung2124; area identified by @annguyenNous in #39590.
