The dashboard MCP catalog only showed name/description/transport and a
non-clickable source. Users couldn't see what an entry connects to or runs
before installing — the exact detail the docs trust model tells them to vet.
- /api/mcp/catalog now returns transport target (url, or command+args),
auth_type, git install source/ref + bootstrap commands, default-enabled
tool hint, and post-install guidance per entry.
- McpPage renders the endpoint URL (http) or command+args (stdio), the git
install source/ref, a collapsible bootstrap-commands list, setup notes,
and the source as a clickable link when it's a URL.
- Docs: drop the 'uv pip install -e .[mcp]' quick-start step (Hermes does
not support pip installs; MCP ships with the standard install) and note
the dashboard now surfaces this detail.
- Strengthen the catalog endpoint test to assert the new inspection fields.
Epic's experimental Unreal MCP plugin embeds an MCP server inside the
Unreal Editor process, served over local HTTP (127.0.0.1:8000/mcp by
default). HTTP transport, no auth, no install block — the user enables
the plugin in-editor and Hermes connects to the URL.
Also drops test_optional_mcps_manifests_ship_in_both_wheel_and_sdist:
it asserted wheel/sdist packaging targets for pip/Homebrew/Nix installs,
which Hermes does not support — installs run from the repo checkout, where
the catalog is discovered by directory iteration with no packaging step.
* fix(tui): don't make Enter swallow trailing-space-only slash completions
Submitting a slash command in the TUI took three Enter presses: one to
complete the name (/ex → /exit), a second that only appended the trailing
space the gateway adds to keep the classic-CLI prompt_toolkit dropdown open
(/exit → "/exit "), and a third to actually submit.
The composer's submit handler accepted the highlighted completion whenever
applying it changed the input at all, so the whitespace-only delta ate an
extra keypress. Treat a completion whose only change is trailing whitespace
on an already-complete token as "already complete" and fall through to
submit. Partial-name and argument completions (a real token change) still
accept on Enter as before.
The replace/accept logic is extracted into pure helpers (applyCompletion,
completionToApplyOnSubmit) in domain/slash.ts.
* test(tui): cover Enter/completion trailing-space behavior and isolate poller queue
- completionApply.test.ts asserts completionToApplyOnSubmit accepts real
token completions (partial command name, argument) but returns null for a
trailing-space-only delta on an already-complete command, so Enter submits
instead of needing extra presses.
- test_notification_poller_delivers_completion / _skips_consumed previously
shared the process-global process_registry.completion_queue. Their events
carry no session_key, so a leaked/concurrent poller could dequeue and
dispatch them to a fixture agent without run_conversation, flaking CI
("AttributeError: '_FakeAgent' object has no attribute 'run_conversation'").
Isolate the queue per test (fresh queue.Queue via monkeypatch), matching the
sibling poller tests that already do this.
The salvaged guard allowed _rmtree_writable(SKILLS_DIR) itself. No call
site ever passes the root — every site passes a skill subdir or its .bak
sibling — so allowing the root only preserves the #48200 footgun (a dest
that collapses to the root wipes every installed skill). Require a strict
strict-child relationship and update the test that documented the
nonexistent 'full reset' capability.
Defense-in-depth fix for the silent wipe of ~/.hermes/ documented in
#48200. A `hermes update --yes` run silently destroyed a user's
.env, MEMORY.md, kanban.db, custom skills, and scripts. Two changes:
1. `_rmtree_writable` in tools/skills_sync.py now refuses to rmtree
anything outside SKILLS_DIR (the HERMES_HOME/skills/ root).
All five call sites pass paths under SKILLS_DIR, so the guard is
a no-op for current code and a loud, recoverable failure for
any future regression (bad path join, malicious bundled
manifest, stale path in scope after an exception).
2. The default `updates.pre_update_backup` flips from false to
true in hermes_cli/config.py. A few minutes of zip per update
is negligible compared to silent total data loss. Still
overridable; --no-backup still works for one-off opt-out.
Five new tests in TestRmtreeWritableScopeGuard (root path,
hermes home, sibling dir, skills root itself, subdir) plus a
flipped `test_default_enabled_creates_backup` in test_backup.py.
178/178 tests pass in the two affected files. Public method
signatures unchanged, no test-stub blast radius.
Closes#48200
Review feedback from egilewski:
1. Remove trailing whitespace from test docstring and mock patches (lines 1430, 1469, 1476, 1482)
2. Expand test coverage: also verify ANTHROPIC_API_KEY is stripped (not just OPENAI_API_KEY)
Changes:
- Remove trailing whitespace from test file
- Add ANTHROPIC_API_KEY to test environment
- Add assertion verifying ANTHROPIC_API_KEY is stripped from cua-driver subprocess env
- Syntax verified: python3 -m py_compile tests/tools/test_computer_use.py ✓
- Use _sanitize_subprocess_env() to filter Hermes-managed credentials
from the cua-driver subprocess environment (issue #37878)
- Prevents credential exfiltration to the third-party cua-driver binary
- Aligns with existing pattern used by browser-tool and other tools
- Add regression test to verify environment sanitization
The cua-driver is a lower-trust MCP subprocess per SECURITY.md §2.3.
Its inherited environment is now scrubbed by default, removing provider
API keys, gateway tokens, and platform credentials that should not leak
to third-party binaries.
Fixes#37878
The TUI banner reported fewer tools than the classic CLI for the same
config (e.g. 32 vs 38) when an MCP server connected slowly. Root cause:
the agent snapshots `agent.tools` once at build time and never re-reads
the registry. `_make_agent` briefly joins the background MCP discovery
thread (`wait_for_mcp_discovery`, ~0.75s) so fast servers land in that
snapshot, but a server slower than the bound — common for an HTTP MCP
server on first connect — lands *after* the agent is built. Its tools are
then absent from both the agent (uncallable until `/reload-mcp`) and the
banner for the whole session.
The classic CLI doesn't hit this because it re-derives
`get_tool_definitions()` at banner render time (which re-waits for
discovery), so it picks the late tools up.
Fix: after a fresh agent is built and its first `session.info` emitted,
if discovery is still in flight, schedule an off-critical-path daemon that
waits for it to finish, then rebuilds the tool snapshot and re-emits
`session.info` — the same rebuild `/reload-mcp` performs, but automatic.
Both the agent's callable tools and the banner count catch up.
Cache safety: the rebuild runs only while the session is still
pre-first-turn (`_user_turn_count`/`_api_call_count` both 0 → nothing
cached to invalidate). Once the user has sent a message we leave the
snapshot frozen rather than break the cached prompt prefix mid-conversation;
late tools then require an explicit `/reload-mcp` (user-consented), exactly
as today. No-op when discovery finished before the agent build, when the
join times out, when the registry was unchanged, or when the session was
swapped/closed while waiting.
Adds entry.mcp_discovery_in_flight() / join_mcp_discovery() accessors and
covers the matrix (added/none/post-turn/timeout/unchanged/replaced) with
unit tests.
Locks the contract that a non-retryable failure (a Cloudflare 403
"managed challenge" page) returns a short, HTML-free `error` field —
guarding the field path where the raw page was dumped to Discord as
~31 messages.
The test drives the standard chat-completions path with a concrete
model so the turn actually reaches `client.chat.completions.create`,
where the mocked 403 is raised. It asserts the create call happened
(guarding against a vacuous pass — an empty model on the Codex
Responses path would otherwise abort on a validation ValueError before
any API call) and that the summarized error includes "403" while
excluding <html> / _cf_chl_opt. The non-retryable abort path is
provider-agnostic; a Cloudflare managed-challenge 403 can surface on
any provider behind Cloudflare.
Source-level guard (install.ps1 only runs on Windows, so there's no Linux CI
runner to execute it): the astral uv install line must be invoked via the call
operator on a resolved host variable, the bare-`powershell` literal that
produced the field-reported "The term 'powershell' is not recognized" must be
gone, and the resolver must be PATH-independent (Get-Process -Id $PID) and
pwsh-aware.
The prior assertion `all("turn1" in k or "turn2" in k for k in keys)` was
weak on two counts: it passes vacuously when keys is empty (a regression
that lost all state would slip through), and after turn 2 finalizes only
turn 1 lingers, so it only ever inspected turn 1 anyway. Replace it with an
exact check that one key survives, it is turn 1, and turn 2 never merged
into it — the real isolation invariant the test name claims.
Scoping the trace key by turn_id (the prior commit) fixed cross-turn
collisions but introduced a slow leak: _finish_trace only pops a key when a
turn ends cleanly (final response has content and no tool calls), so any
turn that is interrupted, ends on a tool call, or has empty final content
now leaves its uniquely-keyed entry in _TRACE_STATE forever. Previously the
constant per-session key was overwritten by the next turn, capping growth at
~1 entry per session.
Add an LRU cap (_MAX_TRACE_STATE) enforced by _evict_stale_locked, called
under _STATE_LOCK immediately before each insert. It evicts the
least-recently-updated entries (using the previously-dead last_updated_at
field) and ends their root span so nothing dangles. Regression test drives
50 non-finalizing turns against a cap of 8 and asserts the dict stays bounded
with the most-recent turns surviving.
The turn- and api-scoped branches each repeated the same
task/session/thread fallback ladder with only the infix differing. Extract
the shared prefix into _scope_prefix so a future scope dimension touches one
ladder instead of three. The legacy branch still returns a bare task_id (not
the task: prefix) for backward compatibility, so it stays separate.
Output key strings are unchanged; a new test pins them across every
task/session/turn/api combination since the keys are matched across hooks
and any drift would silently break trace finalization.
Cleanup pass on the salvage (behavior-preserving):
- diff_bundled_skill now uses the existing _skill_file_list() helper
instead of reimplementing the rglob/is_file/relative_to file-set
enumeration inline (twice).
- Extract _is_tracked_user_modification(origin_hash, user_hash) and use
it in BOTH the sync loop and list_user_modified_bundled_skills() so the
'kept user edit' rule can't drift between the two sites.
- _read_text_for_diff -> _read_for_diff returns (bytes, text); the binary
branch now compares the bytes it already read instead of re-reading
both files from disk.
- Drop the unused 'user_present' key from diff_bundled_skill's return
contract (no consumer or test ever read it).
- test_update_modified_notice: drop the brittle '>= 2 sites' count-floor
so consolidating the two print paths into a shared helper stays a
welcome refactor; keep the per-site 'count notice => discovery hint'
invariant (still mutation-tested).
The PR added helper-level tests for _trace_key but nothing exercised the
keys through the real hooks. This adds TestTurnTraceIsolation, which drives
on_pre_llm_request / on_post_llm_call across two turns of one gateway
session (task_id == session_id, unique turn_id, api_call_count reset per
turn) and asserts each turn opens its own root trace when the first turn
fails to finalize (tool-only final step). This test fails on the pre-fix
code (only one trace opened, turn 2 absorbed into turn 1) and passes with
the scoping fix.
Also pins the turn_id-over-api_request_id key precedence: the turn-scoped
post_llm_call carries no api_request_id, so it must still resolve to the
same key as the request-scoped hooks or finalization breaks.
Salvage follow-up to the cherry-picked feat/test commits:
- W1: the unpack/install update path in main.py printed the
'~ N user-modified (kept)' notice without the new
'hermes skills list-modified' hint that the git-pull path got.
Mirror the hint to both sites so the count is actionable
regardless of which update path runs.
- W2: 'hermes skills diff <name>' (bundled-vs-stock) now shares the
verb with the gateway write-approval 'diff <id>'. The gateway
handler's docstring + truncation message pointed users to
'/skills diff <id>' on the CLI, which now resolves a bundled skill
by that name instead. Point at the pending JSON file and note the
two diff commands are distinct.
- Add an invariant test asserting every 'user-modified (kept)' notice
in main.py carries the discovery hint (guards sibling drift).
Exercises the real sync pipeline (no mocked comparison logic): a pristine
synced skill is not flagged; an edited one is listed and diffed (modified +
added files); an unknown skill returns not-ok; and `reset --restore` clears
the modified state so revert and discovery stay consistent.
Follow-up to #47663 (streaming multipart upload), fixing two issues that
landed with it.
1. Temp file leaked on client disconnect. The streaming upload endpoint's
except chain caught only HTTPException / PermissionError / OSError — all
Exception subclasses. asyncio.CancelledError, raised when a browser aborts
a large upload mid-stream (the exact NS-501 scenario), is a BaseException,
so it bypassed every except clause and reached a finally that only closed
the file handle and never unlinked the temp file. Every aborted large
upload orphaned a partial `.{name}.*.upload` file (up to ~100 MB) in the
target directory. Cleanup now lives in finally, keyed on a `renamed`
success flag, so the temp file is removed on every non-success exit
including BaseException paths. Added test_stream_upload_cleans_temp_on_cancellation,
which fails on the pre-fix code (leaks the temp file) and passes with the fix.
2. python-multipart pinned to ==0.0.27 instead of ==0.0.20. The package was
already resolved at 0.0.27 transitively (via daytona) before #47663; the
explicit ==0.0.20 pin in the [web] extra and the tool.dashboard lazy-install
set downgraded it. Bumped both to ==0.0.27 and regenerated with `uv lock`,
keeping the lockfile coherent. The base dependency stays >=0.0.9,<1.
Resolves conflicts from the OpenViking churn that merged after #32445 was
opened (#48042/#47662 session-switch + write hardening, #47311/#47973):
- plugins/memory/openviking/__init__.py: keep both __init__ field groups
(the PR's _runtime_start_* alongside main's _prefetch_threads/_shutting_down).
- tests/plugins/memory/test_openviking_provider.py: keep BOTH the PR's new
setup-validation tests and main's session-switch/concurrency tests (disjoint
additions to the same region).
Two fixes layered while reconciling (contributor work otherwise preserved):
- Restore the merged tenant-header contract (#22414/#21232). The PR had changed
_VikingClient defaults to '' and made empty account/user OMIT the tenant
headers; main's contract is that empty falls back to 'default' and the
X-OpenViking-Account/User headers are ALWAYS sent (ROOT API keys need them).
Reverted the constructor to 'account or os.environ.get(..., "default")' and
updated the two PR tests that asserted the omit-when-empty behavior.
- Close a secret-file TOCTOU in the setup writers. _write_env_vars and
_write_ovcli_config wrote the api_key/root_api_key file and chmod 0600
AFTERWARD, leaving a world-readable window on newly-created files. Added
_precreate_secret_file() to create with 0600 before any secret bytes land.
* fix(dashboard): stream file uploads via multipart instead of base64 JSON
The dashboard file manager uploaded files (including backup/restore zip
archives) by reading them client-side with FileReader.readAsDataURL and
POSTing a base64 data URL inside a JSON body to /api/files/upload. For a
large backup this (a) inflates the payload ~33%, (b) buffers the whole
file plus its decoded copy in memory, and (c) reliably trips an upstream
proxy body-size/timeout limit, surfacing as a 502 with the upload
appearing to hang indefinitely (NS-501). Dashboard-only hosted users have
no shell fallback to place the archive, so backup restore was unusable.
Add a streaming multipart endpoint POST /api/files/upload-stream
(UploadFile + Form) that reads the request body in 1 MiB chunks straight
to a sibling temp file, enforces the existing 100 MB size cap as it
streams (413 on overflow, before buffering the whole file), and
atomically renames into place so a partial/aborted/over-limit upload
never clobbers an existing file. The frontend api.uploadFile now sends
multipart/form-data (raw bytes, no base64, browser-set boundary) and
FilesPage passes the File object directly; the dead readAsDataUrl helper
is removed. The legacy base64 JSON endpoint stays for backward compat.
FastAPI's UploadFile/Form require python-multipart, which is NOT pulled in
by fastapi itself, so it is added to the base deps, the [web] extra, and
the tool.dashboard lazy-install set (kept in sync).
Validated: 5 new endpoint tests (roundtrip, multi-chunk >1 MiB,
over-limit 413 without clobbering + no temp-file leak, overwrite=false
conflict, forced-root traversal containment); existing base64 tests still
pass; web typecheck + vite build clean; and a real uvicorn server E2E
(5 MB multipart upload -> HTTP 200 in 0.21s, exact byte match) plus a
30 MB TestClient roundtrip confirm constant-memory streaming end to end.
Reported via beta (NS-501).
* build(deps): regenerate uv.lock for python-multipart (NS-501)
CI ran uv lock --check / uv sync --locked which failed because the
python-multipart dependency add was not reflected in uv.lock. Regenerate
the lockfile (resolves to 0.0.20, matching the [web] extra pin) after
merging current main.
Importing a backup wrote every file from the zip over the target home
wholesale. On a hosted instance this clobbered gateway_state.json with the
source machine's last recorded run/desired state — driving the container-boot
reconciler (container_boot._read_desired_state, which only auto-starts a
gateway whose state is "running") off stale/foreign state and leaving the
gateway stuck "starting", disconnected from the Nous portal.
Add _IMPORT_SKIP_NAMES (gateway_state.json, gateway.pid, cron.pid,
gateway.lock, processes.json) and skip them by basename in run_import, so both
the root profile and named profiles preserve the target's own runtime state.
This mirrors what container_boot._STALE_RUNTIME_FILES already sweeps on every
container boot, and protects against older backups that predate the
backup-side exclusions. The import summary reports which files were preserved.
This is the second half of NS-501 (filed separately as NS-508): the upload
502 was fixed in #47663; this fixes the import-breaks-the-instance half.
The gateway half of relay Phase 3. On a MANAGED boot with relay configured and
no secret pinned, the runtime self-provisions its relay credentials IN-PROCESS:
resolve the agent's own Nous access token (resolve_nous_access_token) -> POST
the connector's /relay/provision asserting its own endpoint + route keys ->
set GATEWAY_RELAY_ID/SECRET/DELIVERY_KEY into os.environ so the immediately-
following register_relay_adapter() reads them and dials out authenticated.
No human, no enrollment token, no disk write — the creds live only in process
memory (save_env_value refuses under managed anyway, and keeping the secret off
any volume is the stronger posture). Stateless: process-env creds don't survive
a restart, so a managed container re-provisions every boot; the connector's
rotation window covers a still-connected prior instance. An explicitly-pinned
GATEWAY_RELAY_SECRET is respected (skip). Self-hosted is unchanged: humans keep
using `hermes gateway enroll`.
Endpoint provenance is gateway-asserted (GATEWAY_RELAY_ENDPOINT +
GATEWAY_RELAY_ROUTE_KEYS, env or gateway.relay_* config) — uniform code path
whether the operator sets it (self-hosted) or NAS stamps it (hosted, the only
case NAS knows the public URL). Both absent -> outbound-only provisioning
(credentials, no inbound routes). The connector scopes the asserted endpoint to
the verified tenant, so it stays within the security model.
- gateway/relay/__init__.py: relay_endpoint(), relay_route_keys(),
_provision_url(), _post_provision(), self_provision_if_managed() (never
raises — a provision failure logs and boots without relay auth).
- gateway/run.py: call self_provision_if_managed() immediately before
register_relay_adapter() in the startup path.
Tests: 12 unit (trigger logic, respect-pinned-secret, in-process env wiring,
endpoint+routes vs outbound-only, fail-soft on token/connector failure);
mutation-checked (drop is_managed guard / pinned-secret guard -> tests fail).
Cross-repo live E2E driver lands on the connector side (depends on this).
EXPERIMENTAL: relay auth scheme may change until >=2 Class-1 platforms validate.
Phase 4F (F.1 + F.2 + F.3, agent side). F.4 is the operator-run live smoke
(needs a NAS deployment); recorded in the PR, not code.
F.1 — on_jobs_changed wiring:
- cron/scheduler.py: _notify_provider_jobs_changed() — resolve the active
provider, call on_jobs_changed(), swallow errors. Lives in scheduler.py (not
jobs.py) so the store stays free of provider imports (no import cycle).
- Wired at the consumer surfaces AFTER a successful mutation: the cronjob model
tool (tools/cronjob_tools.py, create/update/remove/pause/resume) — which the
`hermes cron` CLI also routes through — and the REST handlers
(gateway/platforms/api_server.py, same five). Built-in's no-op default = zero
behavior change on the default path. Sleeping-agent direct jobs.json writes
(no tool/CLI/REST) are covered by reconcile-on-wake in start().
F.2 — config: cron.chronos.{portal_url,callback_url,expected_audience,
nas_jwks_url}. All non-secret; the agent holds no scheduler creds and the
outbound provision call reuses the existing Nous token (no token key). Additive
deep-merge key, no version literal.
F.3 — docs:
- docs/chronos-managed-cron-contract.md: authoritative agent↔NAS wire contract
(the three agent-cron endpoints + inbound /api/cron/fire + the 3-hop trust
model + at-most-once/re-arm semantics). This is what the NAS-side agent builds
against.
- cron-internals.md: "Managed cron (Chronos) for scale-to-zero" section.
- cli-commands.md: cron.provider accepts chronos + the cron.chronos.* keys.
- User docs name no scheduler vendor (QStash is a NAS-internal detail).
INVARIANT re-verified: zero qstash/upstash hits across plugins/cron, gateway,
hermes_cli, tools, website/docs (the one remaining repo hit is an unrelated
Context7 MCP comment in tools/mcp_tool.py).
Tests: test_jobs_changed_notify (5) — notify calls provider hook, swallows
errors, built-in harmless, tool create/remove notify. Full cron + chronos +
webhook + config + api_server_jobs suites green (504 in the cron+chronos+webhook
run).
Phase 4E (E.1 + E.2). The inbound side of Chronos: NAS POSTs the agent when a
one-shot fires; the agent verifies a NAS-minted JWT and runs the job.
E.1 — plugins/cron/chronos/verify.py:
- verify_nas_fire_token(token, expected_audience, jwks_or_key, issuer): verifies
signature against the NAS JWKS (RS/ES family; symmetric rejected), aud == this
agent, exp/nbf, iss, and purpose == "cron_fire" (so a general agent JWT can't
be replayed against the fire endpoint). Returns claims or None; never raises.
Crypto delegated to PyJWT[crypto] (already a declared dep) — no hand-rolled
JWT, no new dependency. No key configured → refuse (never unsigned-decode a
security boundary).
- get_fire_verifier(): pluggable indirection so the DQ-4 escape hatch
(direct per-job cron-key) can swap in with no handler change.
E.2 — gateway/platforms/api_server.py:
- POST /api/cron/fire (registered only when _CRON_AVAILABLE). Authenticated by
the NAS-JWT via get_fire_verifier() — NOT API_SERVER_KEY (NAS holds no API
key; this is the only inbound that triggers remote job execution, so it gets
its own purpose-scoped check). Verifier args come from cron.chronos.* config.
401 on bad/missing/forged token. 400 on missing job_id. On success: 202 +
fire_due runs in the background (so a long agent turn never trips NAS's HTTP
timeout); the store CAS claim inside fire_due de-dupes a scheduler retry.
Tests:
- test_chronos_verify (11): REAL RS256 signing — valid→claims, wrong-aud,
missing/wrong purpose, expired, wrong-iss, tampered-signature (attacker key),
no-key-refuse, empty-token, JWKS-URL key resolution, get_fire_verifier.
- test_cron_fire_webhook (5): valid→202+fire, invalid→401+no-fire, missing
token→401, missing job_id→400, and fire path does NOT require API_SERVER_KEY.
api_server regression suites (214) green.
E.3 (NAS endpoints) is a separate cross-repo PR; the wire contract lands next
(docs/chronos-managed-cron-contract.md).
Phase 4D. The first non-default CronScheduler: plugins/cron/chronos/. Inert
unless cron.provider=chronos; resolve_cron_scheduler falls back to the built-in
if unavailable, so cron never loses its trigger.
Files:
- chronos/__init__.py — ChronosCronScheduler + register(ctx).
* is_available(): config-only, NO network (portal_url + callback_url + a
stored Nous access token via get_provider_auth_state). Returns False →
resolver falls back to built-in.
* start(): reconcile() then RETURN — no blocking loop, no 60s wake (DQ-1:
this is what makes scale-to-zero real; the machine wakes only on a
NAS→agent fire).
* _arm_one_shot(job): POST NAS provision {job_id, fire_at, agent_callback_url,
dedup_key=job_id:fire_at}. Agent owns the time → sub-minute fires survive
(no scheduler 1-minute floor).
* reconcile(): converge NAS arms toward jobs.json — arm missing/changed-time,
cancel orphaned, skip paused. Cold process rebuilds from jobs.json +
idempotent dedup_key.
* on_jobs_changed(): reconcile (re-arm/cancel the affected one-shot).
* fire_due(): ABC default (CAS claim + run_one_job) THEN re-arm the next
one-shot. Job gone (one-shot done / repeat-N exhausted) → no re-arm.
- chronos/_nas_client.py — thin HTTP wrapper for provision/cancel/list using
the agent's existing refresh-aware Nous token (resolve_nous_access_token).
Names no scheduler vendor; holds no scheduler creds.
- chronos/plugin.yaml — discovery metadata.
INVARIANT: zero "qstash"/"upstash" hits in plugins/cron, gateway, hermes_cli,
website/docs — the external scheduler is a NAS-internal detail, never named
agent-side.
Tests (13, all NAS mocked, zero network): is_available off-without-config +
on-with-config + makes-no-network; arm payload incl. sub-minute + noop without
next_run; reconcile arms-all / cancels-orphan / skips-paused / skips-already-
armed; fire_due re-arms next / no re-arm when job gone / no re-arm when claim
lost.
Phase 4C. claim_job_for_fire(job_id, *, claim_ttl_seconds=300) in cron/jobs.py:
under the existing _jobs_lock() file lock, claim a job for a single external
fire so that across N gateway replicas exactly ONE wins. Single-machine
deployments always win (unaffected).
Semantics:
- missing / disabled / paused job → False.
- a fresh fire_claim (younger than claim_ttl_seconds) already present → False
(someone else holds it). Stale claim (crashed winner) → overwrite, so a job
is never wedged forever.
- on win: stamp fire_claim={at, by:_machine_id()}; for recurring (cron/interval)
advance next_run_at (mirrors advance_next_run's at-most-once bump so a stale
re-delivery can't re-fire); one-shots keep next_run_at but the fresh claim
blocks a duplicate retry for the same fire.
- mark_job_run now clears fire_claim on completion so a re-armed recurring job
is claimable again next fire.
_machine_id() (HERMES_MACHINE_ID env, else hostname:pid) is attribution-only;
correctness is the file lock + fresh-claim check, not the id.
This is consumed by CronScheduler.fire_due (Phase 4B). tick is untouched — it
still uses advance_next_run, so the built-in single-machine path is unaffected.
Tests (real store, temp HERMES_HOME): claim-once-then-block + next_run advance,
one-shot no-double-claim, unknown→False, paused→False, stale-claim reclaimable,
mark_job_run clears the claim (recurring re-claimable). tests/cron/ 470 passed.
Phase 4B. Three NON-abstract hooks on the CronScheduler ABC, all with
built-in-safe defaults so the built-in inherits them without overriding and
test_abc_growth_stays_additive stays green (required surface still {name,
start}):
- on_jobs_changed(): post-mutation reconcile hook. Built-in no-op.
- fire_due(job_id): claim the job via the store CAS (claim_job_for_fire,
Phase 4C) then run it through the shared run_one_job (Phase 4A). Returns
False if the claim is lost or the job vanished (repeat-N exhausted between
arm and fire). The inbound webhook (Phase 4E) routes here.
- reconcile(): converge the external registry toward jobs.json. Built-in no-op.
fire_due imports claim_job_for_fire/get_job/run_one_job INSIDE the method, so
this commits cleanly before Phase 4C lands claim_job_for_fire (import-time is
unaffected; tests monkeypatch it with raising=False).
Tests: required-surface-unchanged guard, built-in inherits no-op defaults, and
fire_due's three paths (claim+run, lost-claim→no-run, missing-job→no-run).
tests/cron/ green (20 in test_scheduler_provider.py).
Phase 4A. Factor tick's per-job closure (_process_job: execute → save →
deliver → mark) into a module-level run_one_job(job, *, adapters, loop,
verbose) so the external Chronos provider's fire_due (Phase 4D) reuses the
IDENTICAL body — no duplicated correctness. tick's _process_job is now a thin
wrapper calling run_one_job; the pool/in-flight-guard/contextvars dispatch
logic is unchanged.
run_one_job fires ONE given job; it does NOT decide due-ness, claim, or compute
next_run (tick advances next_run_at under the file lock; an external provider
claims via the store CAS in Phase 4C). Pure refactor, no behavior change.
TDD: test_run_one_job.py characterizes the sequence through tick() first
(test_tick_process_job_sequence, passed pre-extraction), then unit-tests the
helper directly: success sequence, [SILENT]→skip delivery, empty-response soft
failure (#8585), failed-job-still-delivers, exception→mark-failed.
Verified: tests/cron/ 459 passed (was 453 + 6 new); tick behavior unchanged.
The install method (docker/git/pip/...) describes the *running binary*, but
detect_install_method() read it from $HERMES_HOME/.install_method — a shared
DATA directory. The Docker docs deliberately bind-mount $HERMES_HOME
(~/.hermes:/opt/data) so config/sessions/memory persist and can be shared with
a host-side Desktop/CLI install.
When a containerized gateway and a host install share one $HERMES_HOME, the
home-scoped stamp is a single slot describing two installs: the published image
stamps 'docker' on every boot, the host install then reads 'docker' and the
in-app updater refuses to run 'hermes update' ("doesn't apply inside the Docker
container"). Reinstalling the Desktop app from the DMG doesn't help because the
contaminated stamp is re-read every time.
Fix (option 1 — code-scoped stamp):
- detect_install_method() reads <install tree>/.install_method first (next to
the running code, immune to the shared data dir). It falls back to the legacy
$HERMES_HOME stamp for back-compat, but IGNORES a 'docker' home stamp when
not actually containerized — so already-poisoned shared homes self-heal.
- stamp_install_method() writes the code-scoped stamp.
- install.sh stamps $INSTALL_DIR instead of $HERMES_HOME.
- Dockerfile bakes 'docker' into /opt/hermes/.install_method at build time
(inside the immutable block); stage2-hook.sh no longer writes the home stamp
and proactively removes a stale 'docker' one to heal existing shared homes.
Genuine containers still resolve to 'docker' (baked stamp, or legacy home stamp
honored when containerized). Unstamped installs in generic containers still fall
through to git/pip (preserves the #34397 fix).
Phase 2 of the pluggable cron-scheduler refactor. Still no call-site changes;
this wires up provider SELECTION with a hard safety net.
Task 2.1: cron.provider config key (hermes_cli/config.py), empty = built-in.
Additive key — deep-merge picks it up into existing configs with no version
bump (verified: load_config() yields the key on a pre-existing config.yaml).
Task 2.2: plugins/cron/__init__.py — discovery machinery cloned near-verbatim
from plugins/memory/__init__.py, retargeted at CronScheduler /
register_cron_scheduler. Bundled (plugins/cron/<name>/) + user
(/plugins/<name>/) dirs, bundled wins collisions. The built-in is
NOT discovered here — it's core, so the fallback can't be removed.
Task 2.3: resolve_cron_scheduler() in cron/scheduler_provider.py — reads
cron.provider and ALWAYS degrades to built-in (missing / unavailable / load
error / typo all fall back with a warning). cron can never be left without a
trigger.
Deviation from plan: the plan's resolver snippet used cfg_get("cron.provider")
(dotted-string form). The real cfg_get signature is cfg_get(cfg, *keys,
default=) — corrected to cfg_get(load_config(), "cron", "provider", default=""),
matching plugins/memory/__init__.py:349. Tests monkeypatch load_config (not
cfg_get) so the real traversal runs.
Tests: default key empty, discovery returns list, unknown load returns None,
and the four resolver paths (empty→builtin, no-section→builtin,
unknown→builtin, unavailable→builtin, available→used). Full tests/cron/: 453
passed; config suite green (additive key, no migration break).
Phase 1 of the pluggable cron-scheduler refactor (Axis B — the trigger).
No call-site changes; this phase only makes the abstraction exist + tested
in isolation.
Task 1.1: cron/scheduler_provider.py — the EXPERIMENTAL CronScheduler ABC.
Required surface is name + start; is_available()/stop() carry safe defaults.
is_available has a no-network invariant. Docstring marks it experimental
until the Chronos provider (Phase 4) validates the shape.
Task 1.2: InProcessCronScheduler wraps the historical 60s ticker loop, calling
cron.scheduler.tick(sync=False) exactly as the raw ticker does. Uses
stop_event.wait(interval) for responsive stop (both raw tickers already do).
Tests: ABC-is-abstract, default-is_available, the InProcess loop drives tick
and stops, stop() no-op, and test_abc_growth_stays_additive (the forward-compat
guard: required abstractmethods must stay exactly {name, start}, so the three
Phase-4 hooks land as NON-abstract additions).
tick() internals in cron/scheduler.py are byte-unchanged (only new file added).
Phase 0 characterization tests still green. Full tests/cron/: 445 passed.
* feat(relay): authenticate the connector⇄gateway WS channel
The relay gateway may be customer-managed and internet-exposed, so the
connector⇄gateway channel is itself authenticated (distinct from the
platform crypto the relay path sheds). Add gateway/relay/auth.py — a
Python port of the connector's HMAC token + delivery-signature schemes
(relayAuthToken.ts / deliverySigning.ts), verified byte-for-byte against
the connector's compiled TypeScript via cross-language test vectors.
Present an Authorization bearer on the /relay WS upgrade keyed by the
per-gateway secret (resolved from GATEWAY_RELAY_ID / GATEWAY_RELAY_SECRET
in env or config). The connector rejects an unauthenticated/invalid/
revoked upgrade with close 4401.
* feat(relay): signed-HTTP inbound delivery receiver
The connector delivers normalized inbound events to a tenant's gateway
over a signed HTTP POST, not the outbound /relay WS: the connector
instance owning a platform socket is generally not the instance a given
gateway dialed out to, so inbound targets a tenant endpoint that may
load-balance across gateway instances.
Add gateway/relay/inbound_receiver.py — verifies x-relay-signature /
x-relay-timestamp over the EXACT raw request bytes (re-serializing would
break the HMAC: JS JSON.stringify is compact, Python json.dumps spaces)
against the per-tenant delivery key verify list within a 300s replay
window, then dispatches messages to handle_message and interrupts to the
interrupt handler. Wire it into the adapter lifecycle (start in connect()
when a delivery key + bind port are configured, tear down in disconnect();
a purely-outbound dev gateway runs without it).
Refine test_relay_sheds_crypto to distinguish PLATFORM crypto (Discord
ed25519, Twilio/WeCom HMAC — still shed) from the connector⇄gateway
CHANNEL auth (intended): auth.py / inbound_receiver.py are exempt from
the platform-symbol scan but still banned from importing platform-crypto
modules, plus a positive guard that auth.py uses only stdlib hmac/hashlib.
* feat(relay): hermes gateway enroll CLI
Add the gateway half of zero-touch enrollment. `hermes gateway enroll`
resolves a fresh Nous Portal access token (the tenant-proving identity),
POSTs {enrollmentToken, gatewayId} to the connector's /relay/enroll, and
persists GATEWAY_RELAY_ID / GATEWAY_RELAY_SECRET / GATEWAY_RELAY_DELIVERY_KEY
to ~/.hermes/.env. The per-gateway secret authenticates the WS upgrade;
the per-tenant delivery key verifies signed inbound deliveries.
Refuses under is_managed() (hosted installs get the secret stamped in by
the orchestrator). Added as an 'enroll' subcommand on the existing
gateway subparser — not a new top-level command.
* docs(relay): inbound is signed HTTP, not WS; document channel auth
Fix the stale contract: §3/§5 said inbound rode the WS socket (single-
instance only, predates the multi-instance socket-ownership + channel-auth
model). Inbound + connector→gateway interrupt are signed HTTP POSTs to the
tenant endpoint. Add §6.1 documenting the two channel-auth schemes (per-
gateway WS-upgrade secret, per-tenant inbound delivery key) and how they
differ from the platform crypto the relay path sheds.
* test(relay): update build_gateway_parser callers for cmd_gateway_enroll
The enroll subcommand added cmd_gateway_enroll as a required keyword-only
arg to build_gateway_parser, but two existing parser-extraction tests still
called it with only cmd_gateway/cmd_proxy — failing CI with TypeError.
Thread the new handler through both call sites and add a test asserting
`gateway enroll` dispatches to cmd_gateway_enroll with its flags parsed.
* fix(docker): supervised gateway uses --replace to take over stale holder
Inside the s6 container image the per-profile gateway service rendered a
bare `hermes gateway run` (no --replace). When a gateway is started
OUTSIDE s6 — a stray shell `hermes gateway run`, an agent action, or the
Open WebUI helper (scripts/setup_open_webui.sh) — it grabs the
per-HERMES_HOME PID lock first. The supervised slot then execs the bare
`gateway run`, hits the "Another gateway instance is already running"
guard, exits non-zero, and s6 restarts it: a restart loop that floods the
log every ~12s and never binds. The container looks up but the gateway is
permanently down, and dashboard-only users (no shell) cannot recover.
Render the supervised run script as `gateway run --replace` so s6 is
authoritative for its slot: it reaps the stale holder via the hardened
takeover path (takeover marker + SIGTERM->SIGKILL-with-confirmation +
scoped-lock cleanup in gateway/run.py) and binds. This matches the
systemd service path, which already builds its argv with --replace
(_build_gateway_argv / 'nohup hermes gateway run --replace'), and the
intent already documented in _maybe_redirect_run_to_s6_supervision. The
existing HERMES_S6_SUPERVISED_CHILD sentinel still prevents the
run->start->run redirect recursion. Each profile is scoped to its own
HERMES_HOME and s6 guarantees one supervised instance per slot, so there
is no legitimate supervised sibling for --replace to clobber.
Reported via beta (NS-505): gateway.log showed PID 17907 'running
(manual process)' with the guard error repeating every ~12s on
v2026.6.5.
Adds a regression test asserting every gateway-run exec line in the
rendered script (default + named profile, both privilege branches)
carries --replace, and updates the existing render-script assertion.
* fix(ci): remove stray .venv symlink committed into repo
The PR's commit accidentally tracked a .venv symlink pointing at the
developer's local venv (mode 120000 -> /home/ben/nous/hermes-agent/.venv).
The CI test/e2e/build jobs run `uv venv` to create .venv and failed with
`failed to create directory .venv: File exists (os error 17)` because the
checkout already contained the symlink. All test shards aborted in <15s
during setup, before any test ran.
Untrack the symlink and add a bare `.venv` entry to .gitignore (the
existing `.venv/` rule only matches a directory, so a symlink slipped
through).
Salvage corrections on top of @XVVH's #44341:
- Make native web_search injection a 1:1 swap for an already-present client
web_search function, NOT an additive grant. The original unconditionally
appended {"type":"web_search"} on every is_xai_responses turn with any
tools, force-enabling Grok server-side search even when the user never
enabled the web toolset (bypassing Hermes web-provider config + tool-trace
plumbing). Now gated on a client web_search actually being present.
- Reconcile grok-composer context to 200000 (merged in #47908) rather than
262144; 200k is xAI's published usable context window for Composer 2.5,
262144 is the /v1/responses input+output budget.
- Update tests to match scoped behavior + add a no-web-toolset guard test.
- AUTHOR_MAP entry for #44341 salvage.
Incomplete-guard (server-side *_call items at in_progress no longer flip
has_incomplete_items) and preflight built-in-tool allowlist kept as-is.
- model_metadata: grok-composer-2.5-fast → 262144 (OAuth slug not in /v1/models)
- codex transport: inject native {"type":"web_search"} for is_xai_responses;
drop client web_search to avoid duplicate-name 400s
- codex adapter: do not treat in-progress server-side *_call items as incomplete
- tests: adapter, transport build_kwargs, model_metadata, oauth recovery
Weak open models (mimo, nemotron-class) that see tool-call XML/JSON sitting in
file contents or tool output get primed and emit their own structured tool
calls mimicking the payload — usually with an empty/whitespace name. Those
calls can't be fuzzy-repaired toward a real tool, so the dispatch loop returns
an error and the model retries. Before this fix, every empty-name error dumped
the full tool catalog back to the model, which fed the priming loop more names
to mimic and inflated context 3-4x across the retry budget.
A blank/whitespace-only tool name now gets a terse anti-priming error that
tells the model in-context tool-call syntax is DATA, with no catalog dump. A
genuinely-wrong-but-nonempty name (a real typo) still gets the full catalog so
the model can self-correct.
Not a sandbox/auth boundary issue: Hermes never parses tool-call text from
content into executable calls (structured tool_calls only; the lone text->call
parser is the Copilot ACP transport and it also rejects empty names). The
reporter's own debug dump confirms the injection never executed.
Behavior-contract test added: empty-name -> terse error, no catalog; nonempty
unknown -> catalog preserved. Exercised end-to-end via run_conversation against
an in-process mock provider.
* feat(mcp): raise default tool-call timeout 120s -> 300s
Port from openai/codex#28234. Long-running MCP tools (web fetches,
sandboxed builds, deep-research servers) routinely exceed 120s, causing
spurious timeout failures. Codex bumped its default MCP tool timeout from
120 to 300 for the same reason.
- _DEFAULT_TOOL_TIMEOUT 120 -> 300 in tools/mcp_tool.py (per-server
'timeout' config override unchanged)
- update test_default_timeout assertion
- document the default in mcp-config-reference.md
* fix(dump): show commit date instead of release date in hermes dump
The version line in `hermes dump` (the top of the /debug report) appended
the package release date in parentheses, which reads like a wall-clock
"generated at" timestamp and confuses support triage. Replace it with the
date the HEAD commit was actually made, resolved live via
`git log -1 --format=%cd --date=short`, kept next to the commit SHA.
On Docker/wheel installs with no .git the date resolves to '' and the
suffix is simply omitted (the baked SHA still identifies the build).
* fix(desktop): resolve electronDist dynamically + self-heal blocked installs
Supersedes the static-path approach (#48081) and the install-step self-heal
(#48082) with a fix that removes the whole failure class instead of chasing each
symptom. Three distinct faults converged into the June desktop-build outage; this
closes all three.
Root cause (the part #48081 left open — "Gap B"):
build.electronDist was a static relative path in apps/desktop/package.json, but
npm workspace hoisting is NOT deterministic — depending on the npm version and
what else is installed, npm nests the workspace-only electron devDep under
apps/desktop/node_modules/electron OR hoists it to the repo root. A static path
matches only one layout, so a clean install intermittently fails with "The
specified electronDist does not exist". #48081 re-pointed the path at the
nested layout (correct today) but electron-builder reads electronDist
STATICALLY, so any future hoist change silently breaks it again — only caught
by a CI invariant, never self-corrected.
Fix:
- scripts/run-electron-builder.cjs: resolve electron the way Node's runtime does
— require.resolve("electron/package.json") walks node_modules from the desktop
project upward and finds electron wherever npm actually put it. The path can
never drift out of sync with the install layout again, on any OS/npm version.
* dist present -> pass -c.electronDist=<abs>/dist so electron-builder reuses
the unpacked runtime (keeps the #38673 fast path that dodges the 26.8.x
missing-binary re-unpack bug).
* dist absent -> omit electronDist; electron-builder fetches Electron itself
via @electron/get honoring electronVersion + ELECTRON_MIRROR.
package.json: builder script now runs the wrapper; the static build.electronDist
is removed (the resolver owns it).
- main.py / install.sh / install.ps1: on a dependency-install failure where the
electron package staged but its dist is missing (electron's install.js
process.exit(1) on a blocked/throttled binary download — #47266/#47917/#48021),
repopulate the dist via electron's downloader (canonical, then npmmirror.com)
and CONTINUE to the build instead of aborting. npm runs postinstall LAST, so
the only casualty is electron/dist; bailing here is what made the pack-time
mirror self-heal unreachable on a blocked network. Hard-fail only when electron
never staged at all (a genuine dependency error).
- The pack-time mirror fallback now retries the build even when the pre-fetch
can't populate the dist: the wrapper lets electron-builder download Electron
itself via the mirror, so the retry is no longer a no-op (it was, when
electronDist was a static path).
The exact 40.10.2 pin (already on main) keeps the third mode — the native
@electron-internal/extract-zip win32 binding that 40.10.3/40.10.4 ship without a
published prebuild — from recurring.
Tests:
- test_desktop_electron_pin.py: replace the static-path-matches-lockfile
invariant with contracts that there is no hardcoded electronDist to drift, the
builder script routes through the resolver, and the resolver uses Node module
resolution + injects -c.electronDist.
- test_gui_command.py: install-failure self-heal continues to build; genuine
(electron-never-staged) install failure still hard-fails; pack retries under
the mirror even when the pre-fetch is blocked.
Salvages/supersedes the overlapping community work in #48003 (sitkarev),
#48012 (omegazheng), #48033 (james47kjv), and #48082.
Co-authored-by: sitkarev <59806492+sitkarev@users.noreply.github.com>
Co-authored-by: omegazheng <zheng@omegasys.eu>
Co-authored-by: james47kjv <220877172+james47kjv@users.noreply.github.com>
* fix(desktop): narrow Electron self-heal to real missing-dist failures
Follow-up on #48091 to remove the remaining misdiagnosis risk from the
installer/build fallback path (#46785 concern): only take the Electron
repair/retry path when Electron's package files are staged and dist is actually
missing/corrupt.
- main.py: add _electron_pkg_staged_missing_dist() and use it to gate install
failure recovery; fail fast for unrelated npm install errors.
- main.py/install.sh/install.ps1: run cache purge + retry only when dist is
missing; do not retry unrelated tsc/vite/build failures under an
Electron-specific narrative.
- install.sh/install.ps1: tighten install-stage self-heal guard to require both
package.json + install.js and missing dist.
- tests: add coverage that install failure hard-fails when Electron dist already
exists, and update retry test to reflect the tightened recovery condition.
Validation:
- Python tests: 64 passed
- install.sh-related tests included in the run
- Real mac build on this machine:
- npm ci at repo root: success
- cd apps/desktop && npm run pack: success
- electron-builder packaged darwin arm64 and used custom unpacked Electron dist
* refactor(desktop): trim electron self-heal helpers and comments
Deduplicate mirror-retry into _try_redownload_electron_dist / shell
counterparts; shorten wrapper and install-script commentary without
changing recovery semantics.
---------
Co-authored-by: sitkarev <59806492+sitkarev@users.noreply.github.com>
Co-authored-by: omegazheng <zheng@omegasys.eu>
Co-authored-by: james47kjv <220877172+james47kjv@users.noreply.github.com>
- test_ws_transport.py: drives WebSocketRelayTransport against a REAL in-process
websockets server (not a mock socket): handshake (hello->descriptor), inbound
frame -> handler, outbound request/response correlation, follow_up routing,
and clean disconnect failing pending waiters. Skips if websockets is absent.
- test_relay_registration.py: rewritten for the config-driven gate — registers
when GATEWAY_RELAY_URL is set / an explicit url is passed / force=True; no-op
without a URL; trailing slash stripped; adapter constructs through the registry.
Full relay suite: 57 passed.
