feat(relay): Phase 4 thread lifecycle — handoff threads, semantic renames, reply_to context, hello command manifest - #71624
Merged
Merged
Conversation
…ames, reply_to context, hello command manifest
- RelayAdapter.create_handoff_thread → one op-gated thread_create op
(Discord channel thread / Telegram forum topic / Slack named seed root);
None fallback contract preserved for the handoff watcher.
- RelayAdapter.rename_thread → thread_rename with only_if_current_name
no-clobber guard on the wire (connector enforces; Telegram guarded
renames fail safe). The native semantic-rename lane
(_is_discord_auto_thread_lane) lights over the relay via the
connector-stamped auto_thread_created/auto_thread_initial_name markers
parsed onto SessionSource in _event_from_wire.
- reply_to {text, author, is_own} wire parse onto the SAME MessageEvent
reply-context fields native adapters populate.
- gateway/relay/command_manifest.py: the gateway-declared slash-command
manifest (native Discord tree mirror) sent on the DISCORD hello; the
connector reconciles Discord's global registration (additive field,
older connectors ignore).
- Contract doc §OutboundAction ops + Phase 4 semantics sections.
- 12 new tests (tests/gateway/relay/test_relay_threads.py); relay suite
266 passed.
Contributor
૮ >ﻌ< ა ci reviewran on 1ec6ff8 all good! |
alt-glitch
added
type/feature
Nemquae
added a commit
to Magi-AGI/hermes-agent
that referenced
this pull request
Jul 27, 2026
* fix(tests): a run that collects nothing can no longer look green
Three foot-guns in the canonical test runner, each of which cost real
debugging time by making an unverified run look verified.
1. Zero collection across the whole run reported success-shaped output.
Per-file rc=5 is rewritten to rc=0 so a platform-gated file (every test
skipped on this OS) doesn't fail the suite — correct, but it also meant a
run where NOTHING was collected anywhere printed
"0 tests passed, 0 failed (100% complete)" and, with no failures
recorded, could exit 0. Now the run-level guard counts every collected
outcome (passed/failed/skipped/errors/xfailed/xpassed): an all-skipped
file still passes, but zero-collected-anywhere prints an explicit
"✗ NO TESTS RAN — this is NOT a pass" block naming the likely causes and
returns 1.
2. A venv without pytest was selected merely for existing. The probe
accepted any directory with bin/activate, so in a checkout/worktree
without a local .venv it picked the RELEASE venv
(~/.hermes/hermes-agent/venv, no pytest). Every file then died with
"No module named pytest" and the run reported 0 tests. Candidates are now
import-checked for pytest — the same guard the HERMES_PYTHON fallback
already applied — and a skipped candidate is named on stderr.
3. Pytest node ids were silently discarded. This runner is file-granular,
so `tests/foo.py::TestBar::test_baz` isn't an existing path: discovery
dropped it and the run ended "No test files to run" while the selector
looked accepted. Node ids are now translated to the FILE plus an inferred
`-k` on the leaf name (parametrized ids reduced to the function name),
with a note explaining the translation. An explicit caller `-k` wins over
the inferred one.
Tests: 4 behavior contracts in tests/test_run_tests_parallel.py. Verified
by sabotage — reverting the runner fails 3 of the 4 (the fourth pins the
pre-existing all-skipped tolerance so fix 1 can't regress it).
* fix(tests): gate WAL-dependent tests on the linked SQLite's real capability
Two tests fail deterministically on main depending only on which SQLite the
test interpreter links — nothing about the code under test. Both are green in
isolation and red in the suite / on an older library, the worst diagnostic
shape.
Root cause A — WAL is not always WAL. Hermes refuses journal_mode=WAL on
SQLite builds carrying the upstream WAL-reset corruption bug (3.7.0–3.51.2,
excluding backports 3.50.7 / 3.44.6) and falls back to DELETE. On such a
build NO -wal sidecar is ever created, so
test_wal_checkpoint_truncates_wal_file asserts on a file that cannot exist.
Invisible locally when the repo .venv and the Hermes managed runtime link
different versions (observed: .venv 3.50.4 → DELETE, runtime 3.53.1 → WAL),
so the same test passes for one interpreter and fails for the other.
- tests/conftest.py: add a `requires_wal` marker plus a
pytest_collection_modifyitems hook that skips such tests when the linked
library will fall back to DELETE. The skip reason names the actual
version so it is diagnosable rather than mysterious.
- pyproject.toml: register the marker.
- test_kanban_db_repair.py: mark the -wal-sidecar test.
Root cause B — process-global warn-once dedup. The WAL-fallback warning is
emitted at most once per (process, db_label). Any earlier test in
test_kanban_db.py that opens a kanban.db consumes that one-shot, so
test_connect_falls_back_to_delete_on_locking_protocol sees zero warnings and
fails — but only as part of the file, never alone.
- test_kanban_db.py: clear both dedup sets in the test that asserts on the
warning, with a comment explaining the isolation trap.
The gate deliberately does NOT import hermes_state. That module computes
DEFAULT_DB_PATH from get_hermes_home() at import time, so importing it during
collection — before the per-test _isolate_hermes_home fixture redirects
HERMES_HOME — permanently caches the developer's REAL ~/.hermes/state.db for
the whole session. The first version of this change did exactly that and made
tests read a live 31,881-session production database (test_console_engine
asserted "Total sessions: 2" and got 31881). The version predicate is
duplicated instead, and tests/test_conftest_wal_gate.py pins the two
implementations in agreement across every documented upstream boundary plus
guards against the import coming back.
Verified: on SQLite 3.50.4 the sidecar test SKIPS naming the version; on
3.53.1 it RUNS and passes, so coverage is not lost where WAL works. Clean
main fails exactly these 2 tests under
`scripts/run_tests.sh tests/hermes_cli/ tests/test_hermes_state.py`
(9926 passed, 2 failed); with this change the same scope is green.
Tests: 726 passed across the two kanban files, test_hermes_state.py, and the
new gate tests.
* feat(desktop): open the session a @session ref names
An agent-written ref rendered as a chip that went nowhere on click. It now
renders as an ordinary inline link — the agent wrote it mid-sentence, so it
should read like one — with the funnel icon leading the resolved title.
Clicking either surface (that link, or the chip in the user's own message)
opens the session as a tab, the way its sidebar row does.
The tile store loads on click rather than at import: the composer's rich
editor pulls this module in, so a static import would boot the profile store
and its REST routing along with every transcript render.
* fix(session-search): stop the agent restating a linked session's title
The old wording ("no need to also spell out the title") left the model free
to write the link on its own line and then repeat the title in the sentence,
showing the user the same session twice. Say plainly that the link IS the
title and belongs mid-sentence as a noun.
* fmt(js): `npm run fix` on merge (#71532)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* fix(desktop): resolve unpacked spawn-helper before chmod (#71171)
* feat(compression): progress-aware timeouts — stop punishing slow summary models
The gateway's pre-agent session-hygiene compression killed the summary
call at a fixed 30s wall-clock deadline (compression.hygiene_timeout_seconds),
regardless of whether the summary model was hung or merely slow. A reasoning
model happily streaming a large summary was cut off mid-generation, the user
got '⚠️ Context compression timed out after 30.0s', and a 300s failure
cooldown left the session oversized — a doom loop for slow-but-healthy
auxiliary models.
Timeouts are now liveness-based instead of wall-clock-based:
- agent/auxiliary_client.py: new thread-local aux_progress_hook. When
installed (only by context compression today), the primary call_llm
attempt streams (stream=True) and aggregates chunks back into a complete
response, ticking the hook per chunk. The configured timeout then acts
per stream read (idle) instead of as a total budget. Providers that
reject streaming fall back to the plain non-streaming call; auth/payment/
rate-limit/transport errors propagate unchanged into the existing
recovery chains. Codex Responses (per SSE event) and Anthropic Messages
(per stream event, via the new create_anthropic_message on_stream_event
callback) tick the same hook from inside their wire adapters.
- agent/conversation_compression.py: CompressionCommitFence gains
touch_progress()/seconds_since_progress(); compress_context() installs
fence.touch_progress as the progress hook around the compress call.
- gateway/run.py: the hygiene wait loop treats hygiene_timeout_seconds as
an INACTIVITY budget — while the fence reports fresh progress the wait
extends, bounded by the new compression.hygiene_total_ceiling_seconds
(default 600s, clamped >= the idle budget) so a degenerate trickle
stream still dies. The timeout warning now says the summary model
produced no output, which is the only case that still triggers it.
- config/docs: hygiene_total_ceiling_seconds added to DEFAULT_CONFIG and
configuration.md; hygiene_timeout_seconds documented as inactivity-based.
Tests: tests/agent/test_aux_progress_streaming.py (hook plumbing, stream
aggregation incl. tool-call deltas and reasoning deltas, rejection
fallback, ceiling kill, fence progress surface); two new gateway tests
prove a slow-but-streaming worker survives past the fixed timeout
(sabotage-verified: fails with the old fixed deadline) and a
forever-trickling worker is still cut off at the ceiling.
* fix(desktop): route live-turn and history actions by the current session, not a stale closure
Redirect/steer, regenerate, restore-checkpoint, edit-message, and
change-cwd read `activeSessionId || activeSessionIdRef.current`, which
prefers the closure-captured prop whenever it is non-null and only falls
back to the ref once the prop is null.
That precedence is backwards. The actions bag is a stable ref that
wiring.tsx mutates in place (Object.assign), and the pane surfaces are
memoized on that stable ref, so a surface does not re-render when the
active session changes and keeps whichever closure was current when it
last rendered. `activeSessionIdRef` is the authority: it is mirrored
during render in use-session-state-cache, and submit.ts /
use-session-actions pin it imperatively mid-flight without touching the
source prop. The prop is stale by design. `cancelRun` in the same file
already reads the ref exclusively and documents exactly this hazard.
User-visible effect after switching chats: a typed correction was
delivered into the previously focused conversation's live turn (the
"session suddenly working on another chat's task" report), and rewinds
truncated the wrong session's transcript — real data loss, since a
truncating resubmit deletes history after the target ordinal. Nothing
crosses over in stored state, which is why a DB/transcript audit of the
affected session comes back clean.
Also fixes the same defect in `changeSessionCwd`, where a stale target
re-anchored another conversation's workspace, pointing that agent's
terminal/file tools at the wrong project. The now-unused
`activeSessionId` option is dropped from useCwdActions rather than left
as a footgun for the next caller.
Not changed, verified not affected: model-edit-submenu reads the runtime
id via `useStore` (live subscription, re-renders on change), and
use-composer-actions guards on `attachedSessionId === activeSessionId`,
so a stale value there only skips a detach instead of writing
cross-session.
Tests: 4 regression cases pin each action to the current session when
the prop and the ref disagree. Verified to fail against the pre-fix code
(all four reported the stale `rt-abc123` instead of the current
session), including a scripted revert of all four sites.
Co-authored-by: Drew Donaldson <49219012+Automata-intelligentsia@users.noreply.github.com>
* fmt(js): `npm run fix` on merge (#71549)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* fix(desktop): avoid stale actions in memoized surfaces
* test(desktop): cover stale steer action binding
* fix(desktop): keep optional action handlers optional through the latest-actions adapters
The adapters wrapped every field in an arrow function, including the
optional ones. That makes an absent handler unconditionally truthy, and
several children gate on a handler's PRESENCE rather than just calling
it:
- onDismissError -> assistant-message.tsx renders the dismiss button
only when defined
- onRestoreToMessage -> thread/index.tsx gates the restore-confirm flow
- onTranscribeAudio -> use-voice-recorder / use-voice-conversation gate
recording on it
- onLoadMoreMessaging / onLoadMoreProfileSessions -> sidebar paging
So the adapter would paint a dead dismiss button and let voice recording
proceed into a no-op transcription path even when the controller had
deliberately left those handlers off.
Wrap an optional field only when it is currently present, and re-read the
latest value inside the wrapper so the stale-closure fix still applies.
Presence is stable for a given actions object (the controller mutates
fields in place rather than toggling a handler between defined and
undefined), while the closure is what churns — which is exactly what the
indirection re-reads.
Adds two regression tests: absent optional handlers stay undefined, and a
present optional handler still late-binds to the latest closure. The
first was verified to fail against the unconditional-wrapper form.
* feat(relay): Phase 3 interactive — native prompt UX (approvals/confirms/clarify) + react ack lifecycle (#71404)
- RelayAdapter.send_exec_approval / send_slash_confirm / send_clarify:
override the base text fallbacks with ONE platform-abstract `prompt` op
(connector renders Discord components / Telegram inline keyboards /
Slack Block Kit / WhatsApp buttons+lists). Option sets mirror the native
adapters exactly (once/session/always/deny with the same
allow_session/allow_permanent/smart_denied gating; once/always/cancel;
choices + Other). Clarify option ids are positional (c0..cN/other) —
choice text is arbitrary UTF-8, callback budgets are 64 bytes.
- Pending-prompt registry: gateway-minted 8-hex prompt ids →
{kind, session_key, extras}; one answer wins, lazy expiry, unanswered
prompts swept opportunistically. Wire timeout_s stays advisory.
- _consume_prompt_response (wired into _on_inbound AND the Discord
passthrough lane): routes answers to the SAME primitives the native
button handlers call — tools.approval.resolve_gateway_approval,
tools.slash_confirm.resolve, tools.clarify_gateway
resolve/mark_awaiting_text — then acks in-channel. Unknown/expired ids
fall through as command-shaped text (typed-reply degradation, the
relay's analog of the native 'approval expired' edit).
- Discord type-3 stub replaced: an hp1:<prompt>:<option> custom_id decodes
to a structured prompt_response (codec mirrored from the connector's
promptCodec); foreign custom_ids keep the legacy best-effort text shape.
- MessageEvent.prompt_response field + ws_transport wire parsing (additive).
- react ack lifecycle: on_processing_start/complete → `react` ops
(👀 → ✅/❌, remove-then-add), op-gated on supported_ops, best-effort by
contract (a react failure never touches the turn).
- Op gating throughout: a connector not advertising `prompt` gets
success=False from send_exec_approval/send_slash_confirm (run.py's text
fallback takes over — same contract as a failed native button send) and
the base numbered-text clarify; `react` silently no-ops.
- docs/relay-connector-contract.md §4: prompt / prompt_response / react
semantics (callback token, budgets, authorization-parity, foreign-id
behavior, per-platform react mappings).
- tests: tests/gateway/relay/test_relay_interactive.py (19) — option-set
rendering + gating matrices, registry consume-once/expiry, resolver
routing for all three kinds (monkeypatched primitives), fall-through
cases, Discord hp1 decode + foreign-id shape, react lifecycle
(success/failure/cancelled), op-gated/best-effort react.
Cross-repo pair: gateway-gateway 'Phase 3 interactive' PR (prompt/react
senders on all four lanes + interaction ingest).
* feat(aux): force streaming for providers that reject non-stream requests
Some OpenAI-compatible endpoints — notably Tencent Copilot
(copilot.tencent.com) — only accept streaming chat requests; any
non-streaming call returns HTTP 400 (code 11101, 'Non-stream chat
request is currently not supported'). The main conversation loop already
streams, so interactive chat works, but every auxiliary task (title
generation, compression, web extraction) used the non-streaming path and
failed on each call.
_provider_requires_stream() detects stream-only endpoints
(copilot.tencent.com built in, plus user-configurable
auxiliary.stream_only_base_urls substring markers in config.yaml).
Matching sync auxiliary calls route through _create_with_progress
(force_stream=True) and async calls through the new
_acreate_with_stream, aggregating the chunk stream — including tool-call
deltas and reasoning deltas — into a complete response via the shared
_ChatStreamAccumulator.
Salvaged from PR #60686 by @kudi88 onto the progress-aware streaming
machinery from #71508, addressing both sweeper-review gaps: the async
path now consumes the stream with 'async for' (awaiting create() and
iterating synchronously raised on AsyncOpenAI streams), and tool-call
deltas are reassembled instead of dropped (MCP passes tools= through
call_llm). Under force_stream there is no silent non-streaming retry —
a stream-only provider rejects those by definition, so the original
error surfaces to the normal recovery chains.
* feat(compression): stream the summary call on every compression path
The progress-hook streaming from #71508 only activated when a
CompressionCommitFence was present (gateway session hygiene). CLI
/compress and in-loop auto-compression still used the plain
non-streaming summary call, where the SDK timeout is inactivity-based —
a byte-trickling provider that keeps the connection alive could outlive
auxiliary.compression.timeout indefinitely (the gap #69192/#41397 were
built to close).
Fenceless compression callers now install a no-op progress hook, which
routes their summary call onto the same streamed path: the configured
timeout acts on inactivity (slow models finish instead of being cut
off mid-generation), and a degenerate trickle stream is bounded by the
streamed total ceiling (max(600s, 4× the task timeout)) instead of
running forever. No config knob needed — the ceiling machinery ships
with the streaming layer and applies uniformly.
Supersedes the opt-in wall-clock deadline approaches in PR #69192
(@JabberELF) and PR #41397: same guarantee (bounded total compression
wall time even while bytes move) without a daemonized watchdog thread
or a new config surface, and without punishing slow-but-healthy models.
* fix(desktop): keep model picker aligned with session state
* fix(skills): stop treating an upstream skill rename as a user deletion
`sync_skills()` keys the bundled manifest by frontmatter name but computes
the destination from the bundled path. When upstream renames or
recategorizes a skill, the manifest key still matches while the new dest
does not exist yet, so the loop fell into its "in manifest but not on
disk" branch and misread the skill as user-deleted: the user's copy was
stranded at the old path forever and never received another update.
Three skills hit this in the July 2026 reorg (computer-use,
evaluating-llms-harness, serving-llms-vllm) — silently frozen at their
pre-rename content on every machine that ran `hermes update`.
Recovery only moves a stale copy when it is byte-identical to the origin
hash recorded the last time sync wrote it, which proves the directory is
ours rather than the user's work. User-modified copies are kept in place
with a warning, hub-installed paths are never touched, and a genuine
deletion (no copy anywhere on disk) is still respected.
- tools/skills_sync.py: add _recover_renamed_skill() plus the
_index_active_skills() / _read_hub_install_paths() indexes; call it
before classification and report moves via a new `relocated` key.
- hermes_cli/main.py: surface relocations in both `hermes update` skill
sync reporting sites.
- tests: 4 cases covering relocate, user-modified preservation,
hub-installed exemption, and genuine-deletion respect.
* fix(desktop): slash commands target the user's chat, not a new session
`/goal status` reported "No active goal" for a goal that was live: the
desktop's slash pipeline resolved its target session differently than the
submit pipeline, so the command ran against a different session than the
chat on screen.
`/goal` state is persisted per-session in SessionDB (`state_meta` key
`goal:<session_id>`). slash.ts resolved with a bare
`hint || activeRef || createBackendSessionForSend()`, so whenever the
runtime binding was momentarily absent — profile swap, reconnect,
orphan-reap, request timeout — it silently MINTED A NEW SESSION and ran
there. submit.ts already handles this case by resuming the routed stored
session on its owning profile (#55578, #67603).
Extract that ladder into one shared resolver both pipelines use, per the
"one resolver owns each policy" rule in apps/desktop/AGENTS.md. This fixes
the whole class, not just `/goal`: every exec/rpc slash command
(`/usage`, `/status`, `/tools`, …) had the same hole.
A targeted durable conversation whose runtime cannot be rebound now
returns null instead of forking the chat — reporting that a command could
not run beats running it against the wrong session.
* fmt(js): `npm run fix` on merge (#71626)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* fix(update): stop WoA integrity gate rejecting ARM64 after IsWow64Process2 HANDLE truncation (#71381)
* fix(update): bind IsWow64Process2 HANDLE and fall back to GetNativeSystemInfo
The #71218 OS-native probe still rejected correct ARM64 Desktop rebuilds on
Windows-on-ARM when ctypes truncated GetCurrentProcess()'s pseudo-handle and
IsWow64Process2 failed with ERROR_INVALID_HANDLE, falling through to the
lying PROCESSOR_ARCHITECTURE=AMD64 value. Type the HANDLE correctly and use
GetNativeSystemInfo before the env-var fallback.
* test(update): cover WoA IsWow64Process2 handle failure and system-info fallback
Pin the residual #71218 shape where IsWow64Process2 returns FALSE, the env
arch lies as AMD64, and GetNativeSystemInfo must still report ARM64 so the
integrity gate accepts a correctly-built ARM64 Hermes.exe.
* fix(desktop): /goal arg stays editable, kickoff queues when busy, slash header stops echoing long args
Four symptoms from the same /goal flow on desktop:
- Typing '/goal <text>' sealed the command into a directive chip on
Space because /goal was registered without args:true, so the goal
prose rendered awkwardly after a pill. The registry row now matches
/personality and /tools: the arg stays editable text.
- The slash status header echoed the ENTIRE invocation ('slash:/goal
<whole goal prose>') in mono, immediately above the backend notice
that repeats the goal text again, and the kickoff user bubble that
repeats it a third time. The header now carries just the command
token (slash:/goal).
- When the session was busy, handleDispatch rendered 'session busy'
and dropped the dispatch message. For /goal that message is the
kickoff prompt, and the backend has ALREADY set the goal by then —
the goal existed but the agent never heard about it, and later turns
looked goal-unaware (#63352). The busy path now queues the kickoff
on the composer queue: it sends on settle and is visible/editable in
the queue panel meanwhile. Falls back to the old message if the
queue rejects the entry.
- A slash command issued on a fresh draft created the backend session
with no preview, so the sidebar row sat as 'Untitled session' —
and when the kickoff was dropped, auto-title never fired either
(it needs a completed user->assistant exchange). ensureSessionId now
seeds the preview with the typed command.
Tests: registry row contract, busy-path queueing (kickoff neither
sends mid-turn nor vanishes), and the header-token assertion.
* fmt(js): `npm run fix` on merge (#71634)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* fmt(js): `npm run fix` on merge (#71638)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* perf(cli): cut `hermes -w` startup from ~14s to ~1.8s by parallelizing + caching the worktree prune
`_prune_stale_worktrees` runs synchronously before the banner on every
`hermes -w` launch and shells out to git several times per candidate
worktree. On a repo with dozens of accumulated worktrees this dominated
startup: measured 18.5s of a 20.6s cold start, 11.6s on warm repeats.
The `git cherry` patch-equivalence probe was the single largest cost
(9.4s of 11.5s across 24 trees). It is also pure waste on repeat runs: a
tree preserved because it holds unpushed work is re-diff-hashed on every
launch, forever, always reaching the same verdict. On this repo 19 of 24
aged trees were unreapable, so ~11s per launch bought zero reaps.
Two changes, both verdict-preserving:
- Split the loop into a stat-only age filter, a parallel read-only
classification phase (thread pool, bounded to min(8, cpu_count)), and
a serial mutation phase. Only reads are concurrent; unlock/remove/
branch -D stay ordered, so log output and removal order are unchanged.
A pool failure falls back to serial rather than blocking startup.
- Memoize `git cherry` verdicts to
$HERMES_HOME/cache/worktree_merge_verdicts.json, keyed on the exact
`(base_sha, head_sha, max_ahead)` range the verdict was computed from.
Because that key is the complete input to the git call, a cache hit is
identical to recomputation by construction: if either ref moves, the
key changes and real git runs again. Bounded to 1000 entries, written
atomically, and a corrupt/hand-edited cache degrades to recomputation.
Measured on a 44-worktree repo (24 aged candidates):
_prune_stale_worktrees before 11.5s after 2.05s cold / 0.42s warm
hermes -w to banner before 13.9s after 1.79s
Work-preservation is unchanged: all 44 worktrees survived, and every
dirty/unpushed/live-locked guard still fires. Verified the 24 real-tree
verdicts are byte-identical across serial, cold-cache, and warm-cache
runs.
Tests: 75/75 in tests/cli/test_worktree.py (67 existing + 8 new). The
new cache tests were sabotage-verified — swapping the exact-sha key for a
naive path-only key makes two of them fail by deleting a worktree that
had gained unmerged work, which is the data-loss case the key prevents.
* fix(image_gen/codex): remove unsupported tool_choice from request payload (#19505)
The chatgpt.com/backend-api/codex backend 400s on every tool_choice
shape for the hosted image_generation tool — it looks up tool_choice as
a function name and never recognizes hosted-tool entries. Removing the
field from _build_responses_payload() lets the host model decide; the
instructions field nudges it toward the tool.
Salvaged from PR #19979 (originally targeted the old
client.responses.stream call, which no longer exists on upstream/main;
the live request now flows through _build_responses_payload + httpx in
_collect_image_b64).
* fix(image-gen): stop reporting the Codex tool_choice 400 as an account limit
The Codex image backend rejected our own request shape for every account, and
we then translated that rejection into "Image generation is not enabled for the
current Codex account. Switch the image provider to OpenAI API key, FAL, or
xAI." — telling every affected user to abandon a provider that had never
actually been tried. That message is why this reads as a setup failure rather
than a bug: the wire error was replaced with a confident, wrong diagnosis.
Removes the classifier and its exception, so any HTTP failure surfaces
verbatim. The paired request-shape fix (previous commit) is what makes the
400 stop happening; this commit makes the next one diagnosable.
Also fixes error-body truncation: bodies were head-truncated at 500 chars, and
Codex error payloads can carry hundreds of bytes of leading metadata, so the
user got a wall of padding and no message. _summarize_error_body() prefers the
parsed error.message and falls back to a truncated raw body.
Docs: drop the unqualified image-to-image claim for the Codex backend and note
that the hosted tool call cannot be forced, so it is best-effort.
Verified E2E against a local fake Codex backend: success path writes a real PNG
with no tool_choice on the wire; the 400 path now returns api_error carrying
"Tool choice 'image_generation' not found in 'tools' parameter" (148 chars)
instead of the entitlement message. Sabotage run confirms all 4 regression
tests fail when the old behavior is restored.
Refs #19505, #49008, #31335.
* fix(sessions): add offline state database recovery
* fix(sessions): point failed in-place repair at offline recovery
A failed `hermes sessions repair` printed "keep state.db and the backup"
and stopped, so a user whose sessions had vanished had no way to discover
that a non-destructive recovery path exists — the reported dead end.
The failure branch now names the next command, read-only step first, seeded
with the backup path it just preserved. Covered by a CLI-surface regression
test that drives the real subprocess against an unrepairable database.
* fix(managed-uv): retry with explicit patch when bare-minor SQLite repair still resolves vulnerable
Fixes #71250.
`hermes update` could not repair the embedded Python runtime's
vulnerable SQLite (WAL-reset bug range 3.7.0-3.51.2, except backports
3.50.7/3.44.6) on installs where `uv python install <minor>` (bare
request, e.g. "3.11") resolves to an older cached/indexed patch that
still links a vulnerable SQLite build, even though a newer
non-vulnerable patch on the same minor line is available and known to
uv. The smoke test correctly rejected the vulnerable candidate every
time, but the provisioner gave up immediately after one attempt,
leaving the repair permanently stuck in the same failing loop on every
`hermes update` run.
Implements option D from the issue (query the index, retry with an
explicit newer patch), which the reporter identified as cleanest:
- New `_list_available_patches()`: runs `uv python list <minor>
--all-versions --only-downloads --output-format json --no-config`,
filters to cpython/default-variant entries (excluding pypy/graalpy),
and returns known patch versions newest-first. Fails safe (returns
[]) on any network/parse error.
- Refactored the single install+find+probe cycle out of
`_install_safe_python_generation()` into `_attempt_install_generation()`,
reusable per attempt with its own generation directory (so a
rejected candidate's files are fully cleaned up before the next
attempt, matching the existing --reinstall semantics).
- `_install_safe_python_generation()` still tries the bare minor-line
request first (preserves the original comment's rationale: for a
given exact patch, python-build-standalone may have no artifact with
fixed SQLite at all). If that resolves vulnerable, it now queries
`_list_available_patches()` and retries with explicit newer patches,
newest-first, bounded to `_MAX_PATCH_RETRIES` (5) attempts -- each
attempt is a real download+install+probe cycle, so the cap keeps
worst-case repair time bounded.
Sanity-checked `_list_available_patches()` against the real `uv`
binary (0.11.7) in this environment: correctly parses live `uv python
list --all-versions --output-format json 3.11` output, returns 14
patches sorted newest-first starting at 3.11.15.
9/9 new tests pass (retry succeeds with a newer patch, exhausts
gracefully when every known patch is vulnerable, empty patch list
degrades to None without crashing, retry count is bounded, plus direct
JSON-parsing unit tests for realistic/malformed/empty uv output);
47/47 in the full tests/hermes_cli/test_managed_uv.py file (including
the 3 pre-existing tests for the original bare-minor success path,
confirming no regression there).
* fix(managed-uv): don't retry patches at or below the installed version
The retry loop skipped only the current version, so on a uv whose download
catalog is stale it walked backwards through older patches. In #71250 the
newest indexed 3.11 is 3.11.14 — the installed version — so all five retries
went to 3.11.13..3.11.9, each a real download+install+probe+delete cycle that
the existing downgrade guard was always going to reject, before failing anyway.
Only newer patches can carry the SQLite fix. Skipping <= current makes the
stale-catalog case cost one attempt instead of six, and still finds 3.11.15
on the first retry when the catalog knows about it.
* feat(relay): Phase 4 thread lifecycle — handoff threads, semantic renames, reply_to context, hello command manifest (#71624)
- RelayAdapter.create_handoff_thread → one op-gated thread_create op
(Discord channel thread / Telegram forum topic / Slack named seed root);
None fallback contract preserved for the handoff watcher.
- RelayAdapter.rename_thread → thread_rename with only_if_current_name
no-clobber guard on the wire (connector enforces; Telegram guarded
renames fail safe). The native semantic-rename lane
(_is_discord_auto_thread_lane) lights over the relay via the
connector-stamped auto_thread_created/auto_thread_initial_name markers
parsed onto SessionSource in _event_from_wire.
- reply_to {text, author, is_own} wire parse onto the SAME MessageEvent
reply-context fields native adapters populate.
- gateway/relay/command_manifest.py: the gateway-declared slash-command
manifest (native Discord tree mirror) sent on the DISCORD hello; the
connector reconciles Discord's global registration (additive field,
older connectors ignore).
- Contract doc §OutboundAction ops + Phase 4 semantics sections.
- 12 new tests (tests/gateway/relay/test_relay_threads.py); relay suite
266 passed.
* fix(acp): pin the session cwd for the turn
An ACP session registered the client's cwd for the *tools*
(`_register_task_cwd` -> `register_task_env_overrides`) but never pinned it
for the *prompt*. `agent/prompt_builder.py` reports
`Current working directory: {resolve_agent_cwd()}`, and `resolve_agent_cwd()`
reads the `_SESSION_CWD` contextvar, which ACP left unset — so it fell back
to `TERMINAL_CWD` / the launch dir.
The system prompt therefore advertised one root (commonly
`~/.hermes/workspace`, or the install tree) while the tools were rooted at the
editor's project. When the model emitted a *relative* path the tools resolved
it correctly; when it emitted an *absolute* path built from the advertised
root, the write landed outside the client's workspace and the turn still
reported success.
Observed in Zed/Buzz-shaped usage: the first prompt in a fresh workspace
creates the file under `~/.hermes/workspace/` and answers "Done." The client's
directory is untouched. Later sessions on the same cwd appear to work once a
session cwd record exists, which makes it look intermittent.
`_run_agent` already calls `set_session_vars(session_key=session_id)` inside
its `contextvars.copy_context()`, and that helper's `cwd` argument exists to
"pin the logical working directory for this context". It simply was not
passed. `gateway/session_context.py` and `tui_gateway/server.py` both already
pass it; ACP was the only surface that did not.
Adds a regression test asserting that the resolved cwd *during the turn* is
the cwd the client passed to `session/new`. It fails without this change
(resolving the install tree instead of the client's project).
* fix(acp): pin the session cwd for slash-command handlers too
Slash commands run on the event-loop thread, outside the per-turn
contextvars.copy_context() that pins the session cwd for the agent call.
/compress reaches agent._build_system_prompt(), whose "Current working
directory" line comes from resolve_agent_cwd() — so an unpinned handler
rebuilt the prompt against the Hermes install tree and PERSISTED it as
the session's cached prompt, re-poisoning every later turn even though
the turn itself is now pinned.
Pin inside a fresh context copy so the write cannot leak into other
concurrent ACP sessions on the shared loop and needs no teardown.
* fix: curator labels bundled skills as agent-created (#64393)
* feat(curator): surface unmanaged skills and add `curator adopt`
`hermes curator status` reported only the skills it manages, staying silent
about curation-eligible skills it can never touch. On a 237-skill library that
meant 112 skills were invisible to every automatic transition with no signal
anywhere — the curator looked broken when it was working as designed.
A skill becomes curator-managed only when `created_by: agent` lands on its
usage record, and only the background review fork writes that marker. Two
populations therefore never qualify: records written before the marker existed
(no key at all, authorship unknowable) and every foreground
`skill_manage(create)` (unset by design — those skills belong to the user).
- skill_usage: `list_unmanaged_skill_names()` / `unmanaged_report()` enumerate
the blind spot, tagging each row with `has_provenance_key` so the two causes
are distinguishable. `adopt_skill()` writes the marker on user declaration
and refuses bundled, hub-installed, external, and protected built-ins.
Adoption never resets the inactivity clock.
- curator CLI: status prints an `unmanaged (no provenance marker)` block on
BOTH the managed and no-managed-skills paths; new `adopt` verb takes names or
`--all-unmanaged`, with `--dry-run` and a confirmation prompt on bulk.
- skills_sync: `_backfill_optional_provenance()` matched candidates by
repo-derived path only, so a skill installed at `mlops/chroma` that upstream
later moved to `mlops/vector-databases/chroma` was skipped forever and
`hermes skills repair-optional` could not fix it. Falls back to an
unambiguous name match, still gated on identical content, and records the
ACTUAL install path.
Provenance stays a declaration, never an inference: a high patch count proves
the agent MAINTAINS a skill, not that it authored one, since Hermes edits
user-written skills on the user's behalf routinely. An "looks agent-made"
heuristic would eventually archive hand-written work.
Validation: 347 targeted tests pass. Each new test verified via sabotage run
(revert the fix, confirm the test goes red) — one initially passed for the
wrong reason because the fixture pinned `prune_builtins` off, masking the guard
under test; fixed to force the shipped default on.
* lint(desktop): ban atom-mirrored refs via no-restricted-syntax eslint rule
A ref synced from a nanostores atom via useEffect lags the atom by one
render. Callbacks that read the ref instead of the atom get stale values —
cancelRun sent session.interrupt to the wrong session (#66485), and
steerPrompt/restoreToMessage/editMessage had closure-priority stale reads.
The rule catches the three shapes of this antipattern:
- useEffect(() => { ref.current = value }, [value])
- useEffect(() => { ref.current = value; ... }, [value])
- useEffect(() => { setMutableRef(ref, value) }, [value])
All 79 existing hits get eslint-disable-next-line with a comment. The
legitimate ref writes (DOM instances, mount flags, request tokens,
prev-value tracking, prop mirrors) stay suppressed; the 11 real bug
sites will be fixed in the next commit so their suppressions can be
removed.
Rule is scoped to apps/desktop only — ui-tui and tests-js don't use
nanostores and don't have this pattern.
* lint(desktop): exempt benign ref writes flagged by the new atom-mirror guard
The guard matches any `.current` write inside a `useEffect`, so it also
flags refs that are not mirroring a reactive value. Seven such sites
landed on main after the rule was written:
- wiring.tsx: one-shot request-seen sentinel
- billing/plans-view.tsx: previous-confirming tracker for focus return
- config-settings.tsx: autosave bookkeeping
- gateway-settings.tsx: monotonic request-sequence counters
- toolset-config-panel.tsx: mount flag + one-shot provider-choice claim
- voice-provider-fields.tsx: one-shot config seed flag
None mirrors an atom/prop/state value, so none can produce the
one-render-lag stale read the rule exists to prevent. Each gets an
eslint-disable with the specific reason rather than widening the rule,
which would reopen the hole for real mirrors.
Verified the guard still catches the real antipattern: a probe file
mirroring `useStore($activeSessionId)` into a ref via useEffect is
reported. eslint over apps/desktop/src is 0 errors, and the 24 remaining
warnings match the pre-change baseline exactly.
* fmt(js): `npm run fix` on merge (#71667)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* fix(desktop): let a slash command be typed anywhere in the prompt
The `/` trigger was anchored strictly at position 0, so the completion
popover only opened when the slash was the first character. Typing
`please run /clean` mid-message produced nothing to pick, which put every
skill out of reach unless the prompt started with it.
Position 0 and mid-message are genuinely two different things, so detect
them separately. At position 0 a slash is a command invocation and keeps
its arg completion (`/personality alic`). After whitespace it's an inline
reference inside prose, so it completes as a single token and never
expands into an arg step. Both stay trailing-anchored, which keeps file
paths (`src/foo/bar`, `look at /usr/local/bin`) from matching.
* fix(desktop): give ⌘T session tabs a live gateway so slash completions load
`useGatewayRequest` only exposed the gateway through a ref that its own
subscription effect fills in, so a component reading it during render saw
null on mount. Session tiles — what ⌘T and the tab-strip "+" open — did
exactly that and passed `gateway={null}` down to ChatBar. The slash
adapter is disabled without a gateway, so a new tab had no completions at
all while ⌘N, which resolves its gateway through a state-driven memo,
worked fine. Nothing re-rendered the tile afterwards unless the connection
state happened to flip, so it never recovered.
Return the `$gateway` atom as a reactive value alongside the ref and use
it wherever the gateway is needed as a render-time value. The ref stays
for `requestGateway`, which wants a stable identity. The model picker,
model visibility, and settings overlays read the same ref during render,
so they're moved over too.
* fix(project-tree): absorb deleted-worktree sessions into the parent home checkout
A linked worktree at <repo>-<suffix> that has been deleted leaves its
sessions with a dangling cwd: the git probe fails and no git_repo_root was
persisted, so the path-only heuristic promoted each one to its own
standalone project. Every abandoned worktree added another phantom entry to
the sidebar, and they accumulate indefinitely.
Recover the parent by trimming one -<segment> at a time off the basename and
returning the first sibling that resolves. A deleted worktree has no checkout
to return to, so its sessions land in the parent's trunk lane rather than a
lane keyed by the dead path.
Live worktrees are unaffected — they resolve through the git probe and keep
their own lane.
* test(project-tree): cover deleted-worktree fold into parent trunk
Assert the dangling-cwd session joins the parent's main lane and that no lane
keyed by the deleted worktree path survives.
* fix(desktop): keep code and diffs out of the tool overflow window
A run of 3+ adjacent tool calls collapses into the 6.75rem
`.tool-group-scroll` window. That is right for status rows, but a patch
or execute_code row's body is a code block the user reads, and the
window squeezed it to a ~2-row viewport behind a fade mask.
The CSS break-out (`:has([data-tool-row][data-tool-open])`) already
lifts the cap once a row is expanded, which is why most tools are safe
to bound. It can't help here: the user has to notice the clipped block
and expand it before the code is legible.
Extend the existing UNBOUNDABLE_TOOLS opt-out to the code-bearing tools
behind an `isUnboundableTool` predicate, deriving the file-edit names
from `isFileEditTool` so a newly supported edit tool can't be exempt in
one place and clipped in the other. `terminal` stays boundable: console
output is a log tail whose last lines are the ones that matter, which
is exactly what the window pins.
Verified in Chrome against the app's compiled CSS: a 3-call run
carrying a 520px diff rendered at 108px before, full height after.
* fix(desktop): publish the profile default reasoning effort to a store
The composer only seeded effort from agent.reasoning_effort when it was
about to reseed the whole selection, which a sticky manual model pick
skips. The configured value was read and discarded, leaving every other
surface with no way to resolve the profile default.
Mirror it into $defaultReasoningEffort on every config load, independent
of the composer reseed, so the manual pick stays sticky without hiding
the default from the surfaces that need it.
* fix(desktop): stop the model picker overriding the configured effort
Selecting a model applied 'preset.effort ?? medium', so a user running
agent.reasoning_effort: high was silently downgraded to medium on every
model switch and every new chat that inherited the pick. The Thinking
toggle and the effort radio group defaulted to the same literal.
Resolve all of them from the profile default instead, falling back to
DEFAULT_REASONING_EFFORT only when config has not loaded.
* fix(desktop): show the real effort on the composer pill
The pill hardcoded 'Med' whenever the surface had no explicit effort, so
a profile configured for high advertised a level the agent would not
use. Take the profile default as a fallback and drop the now-unused
shell.modelMenu.medium key across the locales.
* fix(desktop): name a Cmd+T session from its first message
A Cmd+T tab's session is created unlisted (openNewSessionTile passes
`listed: false`), so it has no $sessions row until its first turn
persists and a refresh surfaces it. For that whole first exchange the
tab strip and the sidebar read "New session".
Cmd+N has no such gap: its session is created per-send, and
createBackendSessionForSend seeds the optimistic row with the user's
text as the preview. A tile never reaches that path — its runtime id is
already bound, so its createBackendSessionForSend seam is a no-op that
returns the existing id.
Seed the row the same way on a tile's first send. The tab strip already
re-syncs on $sessions (watchSessionTiles lists it in `also`), so the tab
and the sidebar both name the session within the first message; the
server's auto-title supersedes the preview when the turn completes.
Guarded so it only ever fires once per session: no-op on empty text and
on an already-listed row, matching across compression by lineage root so
a re-send can't clobber a real title with a raw message preview.
* fix(desktop): scope mid-message references to skills and keep them as chips
Two corrections to the inline slash reference.
Only skills are offered mid-message now. A built-in like `/model` or `/new`
acts on the app, so it reads as nothing useful in the middle of a sentence —
whereas a skill is exactly the thing you want to point at while describing
work. A leading `/` is unchanged and still offers the full command set.
A picked skill also stays a pill in the sent message. The composer already
inserts one, but the message renderer knew nothing about slash references and
flattened it back to raw text on send. It now parses a mid-prose `/skill` into
a chip segment and renders it with the same styling the composer uses. The
submitted text is untouched — the chip still round-trips to the literal
`/clean` the backend expects — so this is presentation only. Path-like tokens
(`/usr/local/bin`) are excluded: unlike the caret-anchored composer trigger,
this scans finished text, so it also has to reject a token that runs on into
a path.
* refactor(desktop): give reasoning effort one owner
The seven effort levels were enumerated four times in four shapes: a
labels map, a radio-option list, a settings value array, and an enum
list for the config field. Each carried its own hardcoded medium
fallback, which is how the picker drifted from the configured default in
the first place.
Collapse them into lib/reasoning-effort, which owns the scale, the
labels, and the resolve/fallback rules. Net -80/+26 in source. Dropping
the effortLabelKey cast also means the i18n keys are now type-checked
against the canonical list instead of asserted into place.
* fix(desktop): clarify inherited profile gateway
* fix(desktop): localize the inherited-gateway copy for Arabic
The ar locale landed after the inheritance-label fix was written, so it
still told users to set a named profile to Local to inherit the default
and fell back to English for the renamed card.
* fix(update): drop the GetNativeSystemInfo probe that lies under WoA emulation
The residual Windows-on-ARM integrity-gate fix added GetNativeSystemInfo as a
second "OS-native" probe behind IsWow64Process2. Microsoft documents the
opposite behavior: "the API GetNativeSystemInfo also returns emulated
processor details when run from an app under emulation."
On the exact hosts the probe was added to rescue -- an x64 hermes-setup.exe
emulated on an ARM64 Surface -- it therefore reports AMD64, making it a
duplicate of the PROCESSOR_ARCHITECTURE rung already below it rather than a
fallback for it. Its test only passed because the kernel32 fake was hand-fed
ARM64, asserting behavior real Windows does not exhibit.
Replace it with GetMachineTypeAttributes, which answers the question the gate
actually asks -- "can this host load a PE of machine X?" -- instead of
inferring it from an architecture name. It is also the only documented API
that reports AMD64-on-ARM64 emulation support. The gate prefers it and falls
back to the existing name-based mapping on pre-Windows-11 hosts.
The load-bearing half of the previous fix (typing GetCurrentProcess as HANDLE
so IsWow64Process2 stops failing ERROR_INVALID_HANDLE) is unchanged.
Co-authored-by: xxxigm <xxxigm@users.noreply.github.com>
Co-authored-by: Teknium <teknium1@users.noreply.github.com>
* fix(tui): offer skill completions for a slash typed mid-message
A slash only opened completions at position 0, so `please run /cle`
suggested nothing. That is correct for execution — commands only run from
the start of a message — but it also removed the ability to reference a
skill inside prose.
Position 0 and mid-message are now detected separately. A leading slash
stays a command invocation with arg completion and the full command set; a
slash after whitespace is an inline skill reference. Only skills are
offered there, since a built-in like /model or /new acts on the app and
reads as nothing useful inside a sentence.
The gateway tags each completion with its kind, derived from the same
skill-command and skill-bundle providers the completer already consumes,
so the TUI filters on data rather than sniffing the display meta glyphs.
applyCompletion keyed its leading-slash check off the start of the input,
which is only the replace point for a position-0 command. It now keys off
the character before compReplace, so an inline pick lands as
`please run /clean` instead of `please run //clean`. The Tab handler
carried its own divergent copy of that logic and now calls the shared
helper.
* fix(tui): keep a referenced skill styled in the sent message
The composer offers a mid-prose /skill as a completion, but the transcript
knew nothing about slash references and rendered the whole user message as
one flat run of text, so the skill lost its accent the moment it was sent.
User messages are now split into plain and reference runs, and the
reference keeps the accent it wore in the composer. The text is unchanged —
concatenating the segments reproduces the input exactly — so this is
presentation only, and the backend still receives the literal /clean.
The scan has to reject a token that continues into a path, since it runs
over finished text rather than anchoring at the caret: /usr/local/bin would
otherwise style as /usr.
* test(tui): cover inline skill references end to end
inlineSlashTrigger and completionRequestForInput are driven through the
mid-message shape, the position-0 shape, and the path cases that must not
claim a slash (/usr/local/bin, src/foo/bar, and/or, 3 /4).
splitSlashSkillRefs asserts the round-trip invariant — the segments always
rejoin to the exact input — rather than freezing a segment list.
The applyCompletion cases reproduce the reported bug rather than restating
the implementation: reverting the fix fails with
`expected 'please run //clean' to be 'please run /clean'`.
* fix(conversation): prevent stale prefix cache on cwd or runtime surface change
* test(conversation): add runtime surface drift test for stored-prompt staleness check
* fix(conversation): use resolve_agent_cwd() and Platform line for stored-prompt staleness check
* test(conversation): add TERMINAL_CWD gateway cwd tests for resolve_agent_cwd staleness check
* fix(conversation): anchor the cwd staleness read to the host-info block
The whole-prompt scan for "Current working directory:" matched USER project
content, not just Hermes' own host-info line. The prompt embeds AGENTS.md /
CLAUDE.md / .cursorrules in the context tier, which sits AFTER the host-info
block, and line_value() takes the LAST match — so any project file containing
a line starting with that label won the scan.
The comparison then ran runtime state against project prose, a mismatch that
never clears: the stored prompt was rejected on EVERY turn, rebuilding the
system prompt each message and destroying the prefix cache for the whole
session. Strictly worse than the staleness the check exists to catch, and
invisible to CI because no test encoded the invariant.
"last match wins" was only ever safe because Model/Provider/Platform live in
the volatile tier at the very END of the prompt. The cwd line is the one field
read from the STABLE tier near the start, so it needs an anchored read:
locate Hermes' own "User home directory:" line and take the working-directory
line that follows it.
Verified against the real builder on a real AIAgent, not a stub:
- stable cwd, clean project -> prompt REUSED (0 rebuilds)
- stable cwd, AGENTS.md naming the
field -> prompt REUSED (was: rebuild/turn)
- drifted cwd -> rebuilt
- drifted cwd, AGENTS.md naming the
NEW cwd -> rebuilt (prose can't mask drift)
Also re-verified no spurious rebuild across all five cwd-resolution paths
(pinned contextvar, TERMINAL_CWD, launch dir, symlinked workspace,
trailing-slash cwd).
The contributor's fixtures omitted the host-info anchor, so they silently
stopped exercising the cwd path once the read was anchored; reshaped them to
match real prompt structure via a _host_block() helper.
* fix(curator): make the autonomous write policy consistent (#67140)
The background write guard decided ownership from `isinstance(usage_rec, dict)`,
so a local skill with NO usage record passed. That successful write called
bump_patch(), which created a `created_by: null` record — and the identical
write was refused from then on. "Allowed exactly once, then never" is a race
with our own bookkeeping, not a policy. Reproduced on main: patch #1 succeeds,
patch #2 with the same arguments is refused.
Option B from the issue. Option A (split `session_review` from
`scheduled_curator` and let the session fork patch user-owned skills it
consulted) would widen autonomous write permission onto skills the user owns
with no user present to consent — wrong direction for a no-user-present actor.
- skill_manager_tool: missing and explicit-null records now resolve
IDENTICALLY, both fail closed. The refusal names the reason and points at
`hermes curator adopt <name>`.
- background_review: both review prompts told the reviewer to patch any skill
consulted in the session and claimed pinned skills could be improved, while
enforcement refused both. Prompts now list pinned, external, and user-owned
skills as protected, and tell the reviewer to RECOMMEND adoption instead of
attempting a write that will be refused.
- skill_usage: document that `created_by` is a curator-management policy flag,
not a provenance claim, and add `is_curator_managed()` so call sites read as
the question they ask. Field name retained — it is on disk in every
`.usage.json` and renaming would strand those records.
- curator CLI: `hermes curator list-unmanaged` itemizes unmanaged skills with
the reason each is unmanaged (completes the #67139 spec).
Foreground writes are untouched: a user-directed edit to a user-owned skill
still works, including on pinned skills.
Sibling tests: 9 failures in test_skill_manager_tool.py were fixtures that
created record-less skills to exercise OTHER guards (consolidation-delete,
read-before-write) and relied on ownership falling through. Fixed at the
fixture, since the real curator only ever operates on managed sediment. One
test asserted the old "manually authored" wording; rewritten to assert the
behavior contract instead of the string.
Validation: 274 targeted tests + all 7 background-review files (60 tests) pass.
E2E on a temp HERMES_HOME (30 checks) covers the flip, foreground writes,
adoption unblocking, pin semantics, prompt/enforcement parity, and the new verb.
Each new test sabotage-verified: revert the fix, confirm it goes red.
Fixes #67140
* fix(desktop): keep the first-run local-start error from being wiped on mount
Clicking "Install Hermes locally" surfaced no error when the bridge was
missing, if the click landed before React drained the passive effect that
reacted to the first bootstrap snapshot. The effect cleared the transient
button state on every change of setupChoice.activeRoot, and the very first
snapshot counts as a change: it moves the root from null to its initial
value. The choice paints as soon as that snapshot commits, so a fast click
produced an error that the effect then cleared before it ever rendered.
The state now records the root it was produced under and is read back only
under that same root, so leaving a phase still discards it but arriving in
one cannot. Deriving it removes the effect rather than adding a guard to it.
Reproduced by observing the DOM with a MutationObserver instead of findBy*,
which only settles on a timer tick — by then the effect has already run and
the window is closed. Reverting the component change fails the new test.
* fmt(js): `npm run fix` on merge (#71706)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* fix(desktop): read a readiness-probe 401 by whether it was credentialed
The boot readiness probe calls the credential-free /api/health, and only a
404 flips it to the /api/status fallback. Both halves are wrong against a
gated gateway.
/api/health landed in ccab46ca4 (2026-07-24), after the v2026.7.20 tag, so
every container pinned to a release lacks the route. On those backends the
dashboard auth gate runs ahead of the SPA catch-all, so an unknown /api/*
path is rejected as unauthenticated rather than 404 — a credential-free
probe can never observe the 404 that the fallback keys on, and boot loops
until the 45s deadline reporting a healthy backend as "did not become
ready". Upgrading the backend is not a workaround for release-pinned
deployments.
Simulating a 0.19.0 backend (both the route and its PUBLIC_API_PATHS entry
removed, since ccab46ca4 added the two together) shows the probe's 401 means
two different things depending on whether credentials were sent:
credential-free: /api/health -> 401 no_cookie, /api/status -> 200
credentialed: /api/health -> 404, /api/sessions -> 200
So the fix is not "fall back on any 401". Uncredentialed, a gate-shaped 401
identifies a missing route and must fall back. Credentialed, a 401/403 is a
rejected session and must fail fast — falling back would hit the public
/api/status, get a 200, and report a dead session as ready, deferring the
no_cookie to the first real API call.
Split the two cases and tag the credentialed rejection as a terminal reauth
error. A generic 401 without the gate shape, plus 429 and 5xx, keep polling
as before.
* fix(desktop): name the readiness-probe auth and password-gateway guard decisions
Two pure seams in native-auth-decisions.ts, following that module's existing
pattern — the value is the test that pins the contract so the god-file call
sites can't drift back to the buggy shape.
resolveReadinessProbeAuth decides how the boot readiness probe
authenticates, delegating to resolveOauthRestAuth for the oauth case rather
than duplicating the bearer-vs-cookie rule.
oauthGuardMayHardFail fixes a second way a gated gateway is misread.
authModeFromStatus maps the gateway's `auth_required: true` onto 'oauth',
but that flag only means the dashboard is GATED — it says nothing about how
you authenticate. A gateway whose providers are all username/password can
satisfy neither of the pre-flight guard's checks by construction:
start_login raises NotImplementedError, /auth/native/authorize rejects
password providers, and its cookies come from a plain password-login POST
rather than the /auth/callback redirect the OAuth partition is primed for.
The guard therefore threw "uses OAuth, but you are not signed in" one line
before a mintGatewayWsTicket that would have succeeded against that very
partition.
The helper returns false only when EVERY advertised provider is
password-based; an unknown or empty list keeps the strict guard, so backends
predating /api/auth/providers are unaffected.
* fix(desktop): latch a confirmed remote reauth failure so the overlay stays clickable
shouldLatchBackendStartFailure deliberately never latches a remote failure:
remote faults are usually transient and must stay retryable without an app
restart. A confirmed reauth rejection is the exception — it cannot self-heal,
because nothing changes until the user signs in again.
Worse, not latching actively prevents the recovery it was protecting. A
non-latching remote boot failure re-runs startHermes on every
getConnection/api call, re-emits running: true, and the boot-failure overlay
(visible = Boolean(boot.error) && !boot.running) hides itself — so the
"Sign in" button flickers out from under the user before it can be clicked.
Add shouldLatchRemoteReauthFailure as a separate predicate rather than
changing the existing one, so transient remote failures keep self-healing.
The two latches are complementary and never fire for the same failure.
* fix(desktop): wire the credentialed readiness probe and reauth latch into boot
Connects the preceding seams to the boot path:
- buildReadinessHealthProbe picks the probe credentials from the connection's
authMode via resolveReadinessProbeAuth, and reports whether the probe is
credentialed so waitForHermesReady can read a 401 correctly. The bearer
goes through fetchJson's `bearer` option — a raw `headers` object is
ignored there, which would have silently probed uncredentialed.
- authMode is threaded into all three waitForHermes call sites: the primary
remote connect, the profile pool backend, and the SSH tunnel (loopback
forward, token auth), so an old backend behind a tunnel gets the same
compatibility path.
- The confirmed reauth failure latches in remoteReauthFailure and
short-circuits startHermes, and is cleared on every recovery path —
resetHermesConnection (soft/hard apply and reconnect), bootstrap reset,
repair, and a CONFIRMED oauth sign-in. A cancelled or closed login window
leaves the latch set so the overlay stays actionable.
- gatewayAuthProviders reads the public /api/auth/providers (cached per base
URL, failures return []) so the oauth pre-flight guard can skip its hard
failure for an all-password gateway while keeping the strict guard
everywhere else.
* chore(contributors): add email mappings for the gated-health-probe salvage
Co-authored-by: HexLab98 <liruixinch@outlook.com>
Co-authored-by: Kevin Yin <182213728+yinkev@users.noreply.github.com>
Co-authored-by: 0301chris <0301chris@gmail.com>
Co-authored-by: Leo Prodz <leo@gtmcore.ai>
Co-authored-by: stephen lopez <stephenlopez2030@gmail.com>
Co-authored-by: diegomarino <diegomarino@users.noreply.github.com>
* fix(skills): fail closed on unknown curator ownership
- require positive agent ownership proof before any background-review mutation
- fail closed when usage record is missing, malformed, or unreadable
- preserve foreground user-directed mutations and valid agent-owned curator flows
- cover patch, edit, delete, write_file, and remove_file end to end
Closes #67073
* fix(tui,desktop): let authored link text outrank the fetched page title
Both chat renderers resolve link titles over the network and render them
in place of the link's text, unconditionally — so they also replaced text
the agent deliberately wrote. `[#71706](url)` rendered as the whole GitHub
page title, and prose labels were swapped out mid-sentence.
The TUI ordered `fetched || label` and always passed the URL to
useLinkTitle. Desktop looked guarded but wasn't: chat markdown hands
authored text to PrettyLink as `fallbackLabel`, not `label`, so
`useLinkTitle(label ? null : target)` never skipped the fetch and
`fetched || label || fallbackLabel` still let the title win.
Treat an authored label as the intent: it wins, and it skips the fetch. A
label that is just the URL still resolves, since `[url](url)` and `<url>`
are bare links wearing mar…
teknium1
added a commit
that referenced
this pull request
Jul 29, 2026
- New page user-guide/messaging/relay.md: what Relay is, enrollment (hermes gateway enroll), capability handshake (media, native approval/clarify, threads, typing), config keys, troubleshooting — derived from gateway/relay/ + gateway_enroll.py (PRs #48147 #48242 #71300 #71363 #71404 #71624 #69721). - desktop.md: artifacts viewer, timeline rail, find-in-page, multi-window, git review/worktrees, multi-terminal + persistence, theme import, rebindable shortcuts, quick entry, context-usage popover, keep-awake, command palette, Hermes Cloud mode, memory graph — all verified against apps/desktop/ source.
teknium1
added a commit
that referenced
this pull request
Jul 29, 2026
- New page user-guide/messaging/relay.md: what Relay is, enrollment (hermes gateway enroll), capability handshake (media, native approval/clarify, threads, typing), config keys, troubleshooting — derived from gateway/relay/ + gateway_enroll.py (PRs #48147 #48242 #71300 #71363 #71404 #71624 #69721). - desktop.md: artifacts viewer, timeline rail, find-in-page, multi-window, git review/worktrees, multi-terminal + persistence, theme import, rebindable shortcuts, quick entry, context-usage popover, keep-awake, command palette, Hermes Cloud mode, memory graph — all verified against apps/desktop/ source.
1 task
1 task
76 tasks
gabrielcosi
pushed a commit
to gabrielcosi/home-ops
that referenced
this pull request
Aug 5, 2026
….7.30 ➔ v2026.8.3) (#253)
This PR contains the following updates:
| Package | Update | Change |
|---|---|---|
| [ghcr.io/gabrielcosi/hermes-agent](https://github.com/NousResearch/hermes-agent) | minor | `v2026.7.30` → `v2026.8.3` |
---
### Release Notes
<details>
<summary>NousResearch/hermes-agent (ghcr.io/gabrielcosi/hermes-agent)</summary>
### [`v2026.8.3`](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.8.3): Hermes Agent v0.20.0 (2026.8.3)
[Compare Source](https://github.com/NousResearch/hermes-agent/compare/v2026.7.30...v2026.8.3)
##### Hermes Agent v0.20.0 (v2026.8.3)
**Release Date:** August 3, 2026
**Since v0.19.0:** \~3,650 commits · \~1,400 merged PRs · \~5,200 files changed · \~559,000 insertions · \~405,000 deletions · **\~1,200 issues closed** · 650+ contributors
> **The Herald Release.** Hermes is the herald of the gods, and this release makes him one in earnest: he **speaks** (real-time conversational voice with streaming TTS, barge-in, on-device wake words, and hands-free control across the CLI, desktop, and every audio-capable gateway platform), he **carries word to other agents** (A2A v1.0), he **announces events to your systems** (signed outbound webhooks), and he **cites his sources** (grounded research with verifiable citations and fact-checking). Around that spine: the desktop app became a platform (artifacts with live preview, a plugin SDK, quick-entry from anywhere, multiple windows), the CLI got a wave of power commands (`!` shell mode, `/init`, `/diff`, `/context`, `/focus`), compression got smarter and gentler, and the tools themselves now recover from their own failures instead of making the model guess. This release rolls up everything from the v0.19.1 infrastructure patch tag — that window is fully documented here.
***
##### ✨ Highlights
- **Talk to Hermes — streaming, conversational voice with barge-in** — Voice mode used to mean: speak, wait for the whole reply to generate, then listen to one long audio file. Now Hermes speaks clause-by-clause as the response streams, you can interrupt it mid-sentence by just talking (it stops, listens, and the model is told you cut in), and busy-aware silence detection means it doesn't talk over you. This works in CLI voice mode, on the desktop, and through gateway adapters. Talking to Hermes finally feels like a conversation, not a voicemail exchange. ([#​69511](https://github.com/NousResearch/hermes-agent/pull/69511), [#​73862](https://github.com/NousResearch/hermes-agent/pull/73862), [#​74223](https://github.com/NousResearch/hermes-agent/pull/74223), [#​74000](https://github.com/NousResearch/hermes-agent/pull/74000), [#​69602](https://github.com/NousResearch/hermes-agent/pull/69602) — [@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife))
- **Wake words and hands-free control** — Say your own open-vocabulary wake phrase ("hey Hermes", or anything you pick) and Hermes starts listening — detection runs on-device, so no audio leaves your machine while it waits. Multi-profile voice routing means different wake words can reach different profiles, and saying "stop" ends the voice chat on every surface without touching the keyboard. Your terminal is now something you can talk to from across the room. ([#​70509](https://github.com/NousResearch/hermes-agent/pull/70509), [#​73106](https://github.com/NousResearch/hermes-agent/pull/73106), [#​73933](https://github.com/NousResearch/hermes-agent/pull/73933) — [@​teknium1](https://github.com/teknium1))
- **Voice on every platform** — Send a voice note to Hermes on WhatsApp, Feishu, DingTalk, LINE, QQ, Photon, or Weixin and it's transcribed and answered; auto-TTS replies are delivered platform-aware (opus where platforms want opus, captions attached correctly). STT is now fully configurable — its own `hermes tools` category, GUI toggles, dashboard dropdowns, unified language resolution so transcripts stop coming back in the wrong language, and OpenAI's gpt-transcribe support. One unified spoken-text preprocessor cleans markdown, code, and URLs out of speech across all TTS providers. ([#​73515](https://github.com/NousResearch/hermes-agent/pull/73515), [#​73508](https://github.com/NousResearch/hermes-agent/pull/73508), [#​73910](https://github.com/NousResearch/hermes-agent/pull/73910), [#​73513](https://github.com/NousResearch/hermes-agent/pull/73513), [#​73067](https://github.com/NousResearch/hermes-agent/pull/73067) — [@​teknium1](https://github.com/teknium1))
- **Research you can trust — grounded citations with fact-checking** — The new `grounded-citations` skill makes Hermes produce research where every claim is backed by a verifiable source: quotes are matched against the actual page text (not hallucinated), citations link to the exact evidence, and a fact-checking mode turns the same machinery on any document or claim you hand it — it tells you what checks out, what doesn't, and what couldn't be verified. If you use Hermes for research, this is the difference between "sounds right" and "provably sourced." ([#​71698](https://github.com/NousResearch/hermes-agent/pull/71698), [#​77104](https://github.com/NousResearch/hermes-agent/pull/77104) — [@​teknium1](https://github.com/teknium1))
- **Outbound webhooks — Hermes pushes events to your systems** — Until now, integrating with Hermes meant polling or listening on a platform. Now Hermes pushes **signed lifecycle events** (session activity, turn completions, tool events) to any HTTP endpoint you register — with HMAC signatures so your receiver can verify authenticity. Wire Hermes into your CI, your home automation, your dashboards, or any service that speaks HTTP, with no polling loop. ([#​69406](https://github.com/NousResearch/hermes-agent/pull/69406) — [@​teknium1](https://github.com/teknium1))
- **The desktop app becomes a platform — artifacts, plugin SDK, quick entry** — Hermes desktop now renders **artifacts**: versioned cards with sandboxed live preview in a right-rail viewer, so generated HTML/apps run safely next to the chat. A real **plugin SDK** landed with Kanban as its founding plugin, `ctx.download` for handing users files, floating pane placement, and multiple GUI windows. A global-hotkey **quick-entry window** captures a thought into any session from anywhere in your OS. The desktop stopped being a chat client and started being a workbench. ([#​72345](https://github.com/NousResearch/hermes-agent/pull/72345), [#​61173](https://github.com/NousResearch/hermes-agent/pull/61173), [#​74413](https://github.com/NousResearch/hermes-agent/pull/74413), [#​72315](https://github.com/NousResearch/hermes-agent/pull/72315), [#​68259](https://github.com/NousResearch/hermes-agent/pull/68259), [#​73143](https://github.com/NousResearch/hermes-agent/pull/73143) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1))
- **Hermes speaks Agent-to-Agent — A2A v1.0** — A new bundled plugin implements the Agent-to-Agent protocol, so Hermes can discover, talk to, and be driven by other A2A-compatible agents. This closes issue [#​514](https://github.com/NousResearch/hermes-agent/issues/514) — one of the oldest open feature requests in the repo. If you're building multi-agent systems with heterogeneous stacks, Hermes now has a standard wire protocol for joining them. ([#​77109](https://github.com/NousResearch/hermes-agent/pull/77109) — [@​teknium1](https://github.com/teknium1))
- **CLI power-user wave** — `!command` runs a shell command instantly without spending a model turn. `/init` scans your project and generates (or updates) an `AGENTS.md`. `/diff` shows staged/all/session changes from any surface, `/context` breaks down exactly what's filling your context window, `/focus` gives you a reduced-output view with hidden-line recovery, and Ctrl+S stashes a half-written prompt into a browsable panel. Plus `hermes import-agent` migrates your Claude Code or Codex CLI setup into Hermes in one command. ([#​72257](https://github.com/NousResearch/hermes-agent/pull/72257), [#​72178](https://github.com/NousResearch/hermes-agent/pull/72178), [#​72240](https://github.com/NousResearch/hermes-agent/pull/72240), [#​72242](https://github.com/NousResearch/hermes-agent/pull/72242), [#​72302](https://github.com/NousResearch/hermes-agent/pull/72302), [#​72262](https://github.com/NousResearch/hermes-agent/pull/72262), [#​72190](https://github.com/NousResearch/hermes-agent/pull/72190) — [@​teknium1](https://github.com/teknium1), several salvaging long-standing community PRs)
- **Correct the agent mid-turn — redirects** — If Hermes is heading the wrong way, you no longer have to `/stop` and re-explain. Type a correction while it works and the active turn is redirected: work in flight is preserved, the original prompt is kept, and the agent course-corrects with your new guidance. Paired with double-ESC draft discard and a composer undo stack, steering feels like editing, not restarting. ([#​63104](https://github.com/NousResearch/hermes-agent/pull/63104), [#​72339](https://github.com/NousResearch/hermes-agent/pull/72339), [#​74736](https://github.com/NousResearch/hermes-agent/pull/74736) — [@​OutThisLife](https://github.com/OutThisLife))
- **Tools that fix themselves** — A sweep of self-recovery upgrades means the agent wastes far fewer turns on tool friction: truncated terminal output spills to a file the agent can read back, `patch` detects already-applied edits and diagnoses whitespace mismatches, `write_file` verifies content on disk, searches that match nothing probe for near-misses and recover, and common failure classes come back with actionable hints. The default tool-calling iteration limit also jumped 90 → 500 — long autonomous runs stopped hitting an artificial wall. ([#​77041](https://github.com/NousResearch/hermes-agent/pull/77041), [#​76998](https://github.com/NousResearch/hermes-agent/pull/76998), [#​77024](https://github.com/NousResearch/hermes-agent/pull/77024), [#​77055](https://github.com/NousResearch/hermes-agent/pull/77055), [#​77011](https://github.com/NousResearch/hermes-agent/pull/77011), [#​76992](https://github.com/NousResearch/hermes-agent/pull/76992), [#​72176](https://github.com/NousResearch/hermes-agent/pull/72176) — [@​teknium1](https://github.com/teknium1))
- **Compression that respects your conversation** — Context compression got a deep overhaul: proactive tool-result pruning for large-window models, per-turn micro-compaction that amortizes the cost instead of one giant pause, a guaranteed N-user-message tail so recent conversation always survives, progress-aware timeouts that stop punishing slow summary models, and ghost-skill defense so a pruned skill can never silently haunt a session. Thresholds are now configurable per-model and in absolute tokens. Long sessions stay coherent and stop stalling. ([#​70254](https://github.com/NousResearch/hermes-agent/pull/70254), [#​75345](https://github.com/NousResearch/hermes-agent/pull/75345), [#​70250](https://github.com/NousResearch/hermes-agent/pull/70250), [#​71508](https://github.com/NousResearch/hermes-agent/pull/71508), [#​70275](https://github.com/NousResearch/hermes-agent/pull/70275) — [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor), salvaging multiple community PRs)
- **Smart approvals grow up** — `hermes approvals suggest` mines your approval history into allowlist proposals, operators can customize the smart-approval policy, a consecutive-denial circuit breaker stops a misbehaving loop cold, and desktop pairing approvals are profile-correct with a proper surface to answer them from. Plus a new approval gate for docker/podman daemon-redirect commands. Less clicking "approve", without giving an inch of control. ([#​72259](https://github.com/NousResearch/hermes-agent/pull/72259), [#​72186](https://github.com/NousResearch/hermes-agent/pull/72186), [#​72203](https://github.com/NousResearch/hermes-agent/pull/72203), [#​74446](https://github.com/NousResearch/hermes-agent/pull/74446), [#​71092](https://github.com/NousResearch/hermes-agent/pull/71092) — [@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife))
- **Faster everywhere, again** — Prompt caching now covers tool schemas on native Anthropic without history loss. `hermes -w` cold start dropped \~14s → \~1.8s, `hermes update` no-ops got 2–6s faster, heavy SDKs lazy-load off the import path, config reads stopped deep-copying (54× faster on the telemetry gate), and the desktop shipped a second 60fps wave — streaming cost independent of transcript length, drag at 60fps with five streaming tabs, idle CPU near zero in the background. ([#​76032](https://github.com/NousResearch/hermes-agent/pull/76032), [#​71637](https://github.com/NousResearch/hermes-agent/pull/71637), [#​74218](https://github.com/NousResearch/hermes-agent/pull/74218), [#​74204](https://github.com/NousResearch/hermes-agent/pull/74204), [#​71835](https://github.com/NousResearch/hermes-agent/pull/71835), [#​72346](https://github.com/NousResearch/hermes-agent/pull/72346), [#​75218](https://github.com/NousResearch/hermes-agent/pull/75218) — [@​kshitijk4poor](https://github.com/kshitijk4poor), [@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife))
- **New places to run and be reached** — Buzz lands as a bundled gateway platform (Block's Nostr-based messenger, with native WebSocket transport and NIP-42 auth), the Vercel AI Gateway provider and Vercel Sandbox terminal backend return modernized, desktop gains an SSH remote-backend connection mode, and the Relay shipped four phases of parity — media, interactive prompts, thread lifecycle, typing indicators — plus HSP personal + org skill sync. ([#​73610](https://github.com/NousResearch/hermes-agent/pull/73610), [#​73761](https://github.com/NousResearch/hermes-agent/pull/73761), [#​74518](https://github.com/NousResearch/hermes-agent/pull/74518), [#​68130](https://github.com/NousResearch/hermes-agent/pull/68130), [#​71300](https://github.com/NousResearch/hermes-agent/pull/71300)–[#​71624](https://github.com/NousResearch/hermes-agent/pull/71624), [#​66730](https://github.com/NousResearch/hermes-agent/pull/66730) — [@​teknium1](https://github.com/teknium1), [@​yoniebans](https://github.com/yoniebans), [@​benbarclay](https://github.com/benbarclay))
***
##### 🎙️ Voice & Speech
##### Conversational voice
- Streaming, conversational TTS with barge-in across all surfaces; clause-by-clause synthesis for CLI voice mode + gateway adapters ([#​69511](https://github.com/NousResearch/hermes-agent/pull/69511), [#​73862](https://github.com/NousResearch/hermes-agent/pull/73862) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1))
- Voice chat UX polish — busy-aware silence, stop hint, thinking sounds, barge-in fix; full-duplex turn listener (interrupt by voice during generation AND playback) ([#​74000](https://github.com/NousResearch/hermes-agent/pull/74000), [#​74223](https://github.com/NousResearch/hermes-agent/pull/74223) — [@​teknium1](https://github.com/teknium1))
- On-device wake words with open-vocabulary phrases + multi-profile voice routing; say "stop" to end voice chat hands-free on every surface ([#​70509](https://github.com/NousResearch/hermes-agent/pull/70509), [#​73106](https://github.com/NousResearch/hermes-agent/pull/73106), [#​73933](https://github.com/NousResearch/hermes-agent/pull/73933) — [@​teknium1](https://github.com/teknium1))
- The model is told when the user interrupts its spoken reply; desktop speaks the whole turn and idle-flushes held narration ([#​69602](https://github.com/NousResearch/hermes-agent/pull/69602), [#​69936](https://github.com/NousResearch/hermes-agent/pull/69936) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1))
- 15-item CLI/TUI voice-mode UX and environment fix wave ([#​73520](https://github.com/NousResearch/hermes-agent/pull/73520) — [@​teknium1](https://github.com/teknium1))
##### TTS / STT infrastructure
- Unified spoken-text preprocessing + speed/instructions/provider tool params; unified STT language resolution (fixes the wrong-language transcription class); global `stt.language` defaults to `en` ([#​73513](https://github.com/NousResearch/hermes-agent/pull/73513), [#​73067](https://github.com/NousResearch/hermes-agent/pull/73067), [#​73100](https://github.com/NousResearch/hermes-agent/pull/73100) — [@​teknium1](https://github.com/teknium1))
- Fully configurable STT — `hermes tools` category, GUI toggle/matrix, dashboard dropdowns, setup status; OpenAI gpt-transcribe support ([#​73910](https://github.com/NousResearch/hermes-agent/pull/73910), [#​73853](https://github.com/NousResearch/hermes-agent/pull/73853) — [@​teknium1](https://github.com/teknium1))
- Platform-aware auto-TTS voice delivery (opus platforms, streamed/global gap, captions); inbound voice classification/routing for Feishu, DingTalk, LINE, QQ, Photon, WhatsApp, Weixin ([#​73508](https://github.com/NousResearch/hermes-agent/pull/73508), [#​73515](https://github.com/NousResearch/hermes-agent/pull/73515) — [@​teknium1](https://github.com/teknium1))
- Command TTS/STT provider hardening — idle timeouts, env scrubbing, no-shell, path guards ([#​73514](https://github.com/NousResearch/hermes-agent/pull/73514) — [@​teknium1](https://github.com/teknium1))
- Sync per-sentence TTS synthesis pipelined with playback — the next sentence renders while the current one speaks ([#​77355](https://github.com/NousResearch/hermes-agent/pull/77355) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Discord voice PCM streams to ffmpeg stdin instead of a temp file ([#​76970](https://github.com/NousResearch/hermes-agent/pull/76970) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
##### 🏗️ Core Agent & Architecture
##### Compression & context
- Proactive tool-result pruning for large-window models; per-turn micro-compaction; N-user tail guarantee (`compression.min_tail_user_messages`); bounded summarizer input with head+tail retention ([#​70254](https://github.com/NousResearch/hermes-agent/pull/70254), [#​75345](https://github.com/NousResearch/hermes-agent/pull/75345), [#​70250](https://github.com/NousResearch/hermes-agent/pull/70250), [#​70249](https://github.com/NousResearch/hermes-agent/pull/70249) — [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Ghost-skill defense — `[SKILL_PRUNED]` markers, protected prune, deterministic survival; progress-aware timeouts; lock-contended compression soft-defers instead of exhausting ([#​70275](https://github.com/NousResearch/hermes-agent/pull/70275), [#​71508](https://github.com/NousResearch/hermes-agent/pull/71508), [#​70285](https://github.com/NousResearch/hermes-agent/pull/70285) — [@​teknium1](https://github.com/teknium1))
- Per-model threshold overrides; absolute token threshold (`compression.threshold_tokens`); opt-in idle-triggered compaction; opt-in progress notices; structured local logging for compression attempts ([#​69339](https://github.com/NousResearch/hermes-agent/pull/69339), [#​69335](https://github.com/NousResearch/hermes-agent/pull/69335), [#​69360](https://github.com/NousResearch/hermes-agent/pull/69360), [#​70457](https://github.com/NousResearch/hermes-agent/pull/70457), [#​69338](https://github.com/NousResearch/hermes-agent/pull/69338) — [@​teknium1](https://github.com/teknium1))
- Context-engine ABC grows `select_context()` + `on_turn_complete()` verbs (salvage of [@​chaos-xxl](https://github.com/chaos-xxl)'s RFC work); engines can suppress or customize compaction status ([#​70458](https://github.com/NousResearch/hermes-agent/pull/70458), [#​69859](https://github.com/NousResearch/hermes-agent/pull/69859) — [@​teknium1](https://github.com/teknium1))
- Strict redaction applied at every compaction text boundary ([#​69294](https://github.com/NousResearch/hermes-agent/pull/69294) — [@​teknium1](https://github.com/teknium1))
##### Prompt caching & hot-path performance
- Tool schemas cached on native Anthropic without history loss + consolidated cache-plan internals ([#​76032](https://github.com/NousResearch/hermes-agent/pull/76032), [#​76067](https://github.com/NousResearch/hermes-agent/pull/76067) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
- DeepSeek prompt caching on OpenCode gateways; per-API-call token accounting off the turn thread; OpenAI wire client reused across sequential LLM calls; send-path tool-call canonicalization memoized ([#​75886](https://github.com/NousResearch/hermes-agent/pull/75886), [#​73359](https://github.com/NousResearch/hermes-agent/pull/73359), [#​73375](https://github.com/NousResearch/hermes-agent/pull/73375), [#​76880](https://github.com/NousResearch/hermes-agent/pull/76880) — [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Readonly config loader at 29 call sites (28× cheaper reads); per-turn config deepcopies killed (telemetry gate 54×); one raw config.yaml parse per process; inter-tool delay removed ([#​74322](https://github.com/NousResearch/hermes-agent/pull/74322), [#​74211](https://github.com/NousResearch/hermes-agent/pull/74211), [#​74228](https://github.com/NousResearch/hermes-agent/pull/74228), [#​64172](https://github.com/NousResearch/hermes-agent/pull/64172) — [@​teknium1](https://github.com/teknium1), [@​Soju06](https://github.com/Soju06))
- Lazy heavy-SDK imports (−8-10% import cost on top of the mcp/tool-discovery diet); streaming hot loop drops per-chunk repr() (\~3× cheaper accounting); cursor/memo optimizations for per-iteration history walks ([#​74204](https://github.com/NousResearch/hermes-agent/pull/74204), [#​74194](https://github.com/NousResearch/hermes-agent/pull/74194), [#​74221](https://github.com/NousResearch/hermes-agent/pull/74221), [#​74231](https://github.com/NousResearch/hermes-agent/pull/74231) — [@​teknium1](https://github.com/teknium1))
- Cold-start \~14s GIL stall during backend init mitigated; turn flush batched into one SQLite transaction; provider-capability-gated prompt cache keys (implied for api.openai.com) ([#​77814](https://github.com/NousResearch/hermes-agent/pull/77814), [#​77619](https://github.com/NousResearch/hermes-agent/pull/77619), [#​77609](https://github.com/NousResearch/hermes-agent/pull/77609) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
- AIAgent hot-path salvage — prompt-cache copy, reasoning-timeout precompute, lazy compressor init ([#​57229](https://github.com/NousResearch/hermes-agent/pull/57229) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
##### Approvals & the agent loop
- `hermes approvals suggest` mines approval history into allowlist proposals; operator-customizable `approvals.smart_policy`; consecutive-denial circuit breaker; cross-surface approvals mode command ([#​72259](https://github.com/NousResearch/hermes-agent/pull/72259), [#​72186](https://github.com/NousResearch/hermes-agent/pull/72186), [#​72203](https://github.com/NousResearch/hermes-agent/pull/72203), [#​63517](https://github.com/NousResearch/hermes-agent/pull/63517) — [@​teknium1](https://github.com/teknium1))
- Docker/podman daemon-redirect commands require approval; session-wide runaway-loop caps for web\_search + delegate\_task (Claude Code-inspired) ([#​71092](https://github.com/NousResearch/hermes-agent/pull/71092), [#​66600](https://github.com/NousResearch/hermes-agent/pull/66600) — [@​teknium1](https://github.com/teknium1))
- Mid-turn redirects — user corrections steer the active turn, preserving in-flight work and the original prompt ([#​63104](https://github.com/NousResearch/hermes-agent/pull/63104), [#​72339](https://github.com/NousResearch/hermes-agent/pull/72339) — [@​OutThisLife](https://github.com/OutThisLife))
- Delegation: structured timeout/stall metadata + live per-child status in `/agents`; subagents can use `execute_code`; redacted child tool history exposed in `subagent_stop`; public subagent lifecycle API for plugins ([#​72300](https://github.com/NousResearch/hermes-agent/pull/72300), [#​69325](https://github.com/NousResearch/hermes-agent/pull/69325), [#​72403](https://github.com/NousResearch/hermes-agent/pull/72403), [#​72501](https://github.com/NousResearch/hermes-agent/pull/72501) — [@​teknium1](https://github.com/teknium1))
- Single-owner refactors for backend identity + failure-scoped skips, empty-content wire repair, call\_id/reasoning sanitization, model-switch parsing ([#​72505](https://github.com/NousResearch/hermes-agent/pull/72505), [#​73071](https://github.com/NousResearch/hermes-agent/pull/73071), [#​74319](https://github.com/NousResearch/hermes-agent/pull/74319), [#​74229](https://github.com/NousResearch/hermes-agent/pull/74229) — [@​teknium1](https://github.com/teknium1))
- Labeled reasoning excerpt surfaced at the empty-response terminal; tool\_search probe-validates blind tool\_call args ([#​65144](https://github.com/NousResearch/hermes-agent/pull/65144), [#​59267](https://github.com/NousResearch/hermes-agent/pull/59267) — [@​teknium1](https://github.com/teknium1))
##### Tool self-recovery wave
- Terminal: recoverable truncation (full output spilled + pre-truncation size), cwd echoed when a command changes directory, output-pattern failure hints ([#​77041](https://github.com/NousResearch/hermes-agent/pull/77041), [#​77004](https://github.com/NousResearch/hermes-agent/pull/77004), [#​76992](https://github.com/NousResearch/hermes-agent/pull/76992) — [@​teknium1](https://github.com/teknium1))
- Patch: already-applied edits return success no-op, whitespace-visualized no-match diagnosis, ambiguous-match locations listed ([#​76998](https://github.com/NousResearch/hermes-agent/pull/76998), [#​77024](https://github.com/NousResearch/hermes-agent/pull/77024), [#​77001](https://github.com/NousResearch/hermes-agent/pull/77001) — [@​teknium1](https://github.com/teknium1))
- Search: zero-match probes + multi-path recovery, auto-multiline for newline patterns; read\_file default limit 500 → 2000 lines; negative-result cache for read/search misses; write\_file verifies on-disk content ([#​77011](https://github.com/NousResearch/hermes-agent/pull/77011), [#​77102](https://github.com/NousResearch/hermes-agent/pull/77102), [#​76996](https://github.com/NousResearch/hermes-agent/pull/76996), [#​76945](https://github.com/NousResearch/hermes-agent/pull/76945), [#​77055](https://github.com/NousResearch/hermes-agent/pull/77055) — [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor))
- execute\_code recovery hints; skill\_view dedup stub for unchanged re-reads; terminal/execute\_code schema prose trimmed \~40%; tiered tool disclosure scales with catalog size; default iteration limit 90 → 500 ([#​77106](https://github.com/NousResearch/hermes-agent/pull/77106), [#​77095](https://github.com/NousResearch/hermes-agent/pull/77095), [#​77023](https://github.com/NousResearch/hermes-agent/pull/77023), [#​67034](https://github.com/NousResearch/hermes-agent/pull/67034), [#​72176](https://github.com/NousResearch/hermes-agent/pull/72176) — [@​teknium1](https://github.com/teknium1))
##### Providers & models
- Vercel AI Gateway provider + Vercel Sandbox terminal backend return, modernized (SDK 0.7.2, telemetry off) ([#​74518](https://github.com/NousResearch/hermes-agent/pull/74518) — [@​teknium1](https://github.com/teknium1))
- Gemini 3.1 Pro + 3.6 Flash in catalogs; Gemini salvage cluster (3.6-flash aux default, Vertex catalog, direct cost tracking); claude-opus-5 in OpenRouter + Nous Portal; deepseek-v4-flash-0731 ([#​73479](https://github.com/NousResearch/hermes-agent/pull/73479), [#​73516](https://github.com/NousResearch/hermes-agent/pull/73516), [#​70946](https://github.com/NousResearch/hermes-agent/pull/70946), [#​75501](https://github.com/NousResearch/hermes-agent/pull/75501) — [@​teknium1](https://github.com/teknium1))
- Bedrock Converse API prompt caching (cachePoint) ([#​70231](https://github.com/NousResearch/hermes-agent/pull/70231) — [@​JoaoMarcos44](https://github.com/JoaoMarcos44))
- OpenAI data-residency endpoints get declared transport + correct catalog; provider-aware API-server request routing; backend-acknowledged session model lock; Nous sticky routing via top-level session\_id ([#​74958](https://github.com/NousResearch/hermes-agent/pull/74958), [#​70853](https://github.com/NousResearch/hermes-agent/pull/70853), [#​70950](https://github.com/NousResearch/hermes-agent/pull/70950), [#​69253](https://github.com/NousResearch/hermes-agent/pull/69253) — [@​victor-kyriazakos](https://github.com/victor-kyriazakos), [@​teknium1](https://github.com/teknium1))
- Model picker: curated defaults + collapsible providers + select-all; stale caches served instantly with background refresh; custom-endpoint probe capped at 1.5s; honcho OAuth device-code login ([#​73172](https://github.com/NousResearch/hermes-agent/pull/73172), [#​76430](https://github.com/NousResearch/hermes-agent/pull/76430), [#​76922](https://github.com/NousResearch/hermes-agent/pull/76922), [#​61608](https://github.com/NousResearch/hermes-agent/pull/61608) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor), [@​akattelu](https://github.com/akattelu))
- ACP: named custom providers in the model selector; authenticated cross-provider model choices; non-blocking startup via background MCP discovery ([#​70082](https://github.com/NousResearch/hermes-agent/pull/70082), [#​70404](https://github.com/NousResearch/hermes-agent/pull/70404), [#​75985](https://github.com/NousResearch/hermes-agent/pull/75985) — [@​israellot](https://github.com/israellot), [@​amanning3390](https://github.com/amanning3390), [@​kshitijk4poor](https://github.com/kshitijk4poor))
##### Secrets & config
- Command-helper secret source (composes with all vaults); one-command token rotation + actionable startup errors; opt-in encrypted break-glass cache for Bitwarden; vault-injected keys scoped per profile home; orchestrator preserve\_existing + profile aliasing ([#​69266](https://github.com/NousResearch/hermes-agent/pull/69266), [#​68605](https://github.com/NousResearch/hermes-agent/pull/68605), [#​69251](https://github.com/NousResearch/hermes-agent/pull/69251), [#​69250](https://github.com/NousResearch/hermes-agent/pull/69250), [#​69058](https://github.com/NousResearch/hermes-agent/pull/69058) — [@​teknium1](https://github.com/teknium1))
- `${env:VAR}` SecretRef parity between config.yaml and MCP config; secret-source env vars reach stdio MCP servers ([#​69267](https://github.com/NousResearch/hermes-agent/pull/69267), [#​69053](https://github.com/NousResearch/hermes-agent/pull/69053) — [@​teknium1](https://github.com/teknium1))
- Canonical config loaders for behavioral reads; table-driven config migration registry; DEFAULT\_CONFIG extracted to config\_defaults.py; auto-migration support floor at v12 ([#​74237](https://github.com/NousResearch/hermes-agent/pull/74237), [#​74200](https://github.com/NousResearch/hermes-agent/pull/74200), [#​74182](https://github.com/NousResearch/hermes-agent/pull/74182), [#​74433](https://github.com/NousResearch/hermes-agent/pull/74433) — [@​teknium1](https://github.com/teknium1))
##### 🌐 Gateway, Relay & Fleet
- Session activity heartbeats, stall watchdog, and bounded compression waits — re-landed hardened after an in-window revert cycle (originally [#​72424](https://github.com/NousResearch/hermes-agent/issues/72424) by [@​fangliquanflq](https://github.com/fangliquanflq)) ([#​76354](https://github.com/NousResearch/hermes-agent/pull/76354) — [@​teknium1](https://github.com/teknium1))
- SessionState consolidation (19 session-keyed dicts → one turn/conversation/persistent-scoped object); TurnContext/TurnRunner seam extraction; declarative busy\_policy on CommandDef ([#​74289](https://github.com/NousResearch/hermes-agent/pull/74289), [#​74353](https://github.com/NousResearch/hermes-agent/pull/74353), [#​74197](https://github.com/NousResearch/hermes-agent/pull/74197) — [@​teknium1](https://github.com/teknium1))
- Relay parity waves: Phase 1 (supported\_ops discovery, identity fields, /handoff aliasing), Phase 2 media, Phase 3 interactive prompts, Phase 4 thread lifecycle; egress typing indicators ([#​71300](https://github.com/NousResearch/hermes-agent/pull/71300), [#​71363](https://github.com/NousResearch/hermes-agent/pull/71363), [#​71404](https://github.com/NousResearch/hermes-agent/pull/71404), [#​71624](https://github.com/NousResearch/hermes-agent/pull/71624), [#​69721](https://github.com/NousResearch/hermes-agent/pull/69721) — [@​benbarclay](https://github.com/benbarclay))
- HSP skill sync: personal client (M1) + org-skills client (M2) + org-skill namespace with token-gated discovery ([#​66730](https://github.com/NousResearch/hermes-agent/pull/66730), [#​70024](https://github.com/NousResearch/hermes-agent/pull/70024), [#​70459](https://github.com/NousResearch/hermes-agent/pull/70459) — [@​benbarclay](https://github.com/benbarclay))
- Buzz (Block/Nostr) platform adapter with native WebSocket inbound transport + NIP-42 auth ([#​73610](https://github.com/NousResearch/hermes-agent/pull/73610), [#​73761](https://github.com/NousResearch/hermes-agent/pull/73761) — [@​teknium1](https://github.com/teknium1))
- Photon: native polls, effects, clarify-as-poll, rich links (4-PR salvage) ([#​73614](https://github.com/NousResearch/hermes-agent/pull/73614) — [@​teknium1](https://github.com/teknium1))
- Slack: native Block Kit clarify buttons; opt-in reaction triggers; outbound payload sanitization; thread-context lifecycle fixes ([#​69318](https://github.com/NousResearch/hermes-agent/pull/69318), [#​70195](https://github.com/NousResearch/hermes-agent/pull/70195), [#​69317](https://github.com/NousResearch/hermes-agent/pull/69317), [#​69320](https://github.com/NousResearch/hermes-agent/pull/69320) — [@​teknium1](https://github.com/teknium1))
- Discord auto-thread sessions keyed on prospective\_thread\_id; reply references built from ids (no fetch\_message); WhatsApp configurable inbound read receipts ([#​76513](https://github.com/NousResearch/hermes-agent/pull/76513), [#​76875](https://github.com/NousResearch/hermes-agent/pull/76875), [#​73322](https://github.com/NousResearch/hermes-agent/pull/73322) — [@​benbarclay](https://github.com/benbarclay), [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Kanban wakes resume the creator's DM/thread session; kanban/delegate wake-ups reach api\_server sessions; per-task model + thinking-depth from the board ([#​72191](https://github.com/NousResearch/hermes-agent/pull/72191), [#​70171](https://github.com/NousResearch/hermes-agent/pull/70171), [#​69876](https://github.com/NousResearch/hermes-agent/pull/69876), [#​76417](https://github.com/NousResearch/hermes-agent/pull/76417) — [@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife))
- Relay: Discord tool-progress routed into the auto-thread instead of the parent channel ([#​77830](https://github.com/NousResearch/hermes-agent/pull/77830) — [@​benbarclay](https://github.com/benbarclay))
- Outbound webhooks — push signed lifecycle events to external endpoints; simplex channel enumeration in `hermes send --list` ([#​69406](https://github.com/NousResearch/hermes-agent/pull/69406), [#​77110](https://github.com/NousResearch/hermes-agent/pull/77110) — [@​teknium1](https://github.com/teknium1))
##### 🖥️ Hermes Desktop App
##### The platform wave
- **Artifacts** — versioned cards, sandboxed live preview, right-rail viewer ([#​72345](https://github.com/NousResearch/hermes-agent/pull/72345) — [@​teknium1](https://github.com/teknium1))
- **Plugin SDK** — Kanban as the founding desktop plugin; `ctx.download` hands the user a file; widget-app SDK (apps as state+reducer+render) with three reference apps; widget-grid layout engine + background-aware theme engine ([#​61173](https://github.com/NousResearch/hermes-agent/pull/61173), [#​74413](https://github.com/NousResearch/hermes-agent/pull/74413), [#​68306](https://github.com/NousResearch/hermes-agent/pull/68306), [#​20379](https://github.com/NousResearch/hermes-agent/pull/20379) — [@​OutThisLife](https://github.com/OutThisLife))
- Quick-entry window (global hotkey → any session); multiple GUI windows; floating pane placement; pane toggles anywhere + hidden header; ⌘O open-folder-as-project ([#​72315](https://github.com/NousResearch/hermes-agent/pull/72315), [#​68259](https://github.com/NousResearch/hermes-agent/pull/68259), [#​73143](https://github.com/NousResearch/hermes-agent/pull/73143), [#​75848](https://github.com/NousResearch/hermes-agent/pull/75848), [#​74623](https://github.com/NousResearch/hermes-agent/pull/74623) — [@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife))
- SSH remote-backend connection mode; event-driven live sync replaces always-on polls; remote profile routing/sessions/pool lifecycle repaired ([#​68130](https://github.com/NousResearch/hermes-agent/pull/68130), [#​73673](https://github.com/NousResearch/hermes-agent/pull/73673), [#​72835](https://github.com/NousResearch/hermes-agent/pull/72835) — [@​yoniebans](https://github.com/yoniebans), [@​OutThisLife](https://github.com/OutThisLife))
- Let the agent drive the shell (preview pane + pane focus) AND inspect the desktop app it's developing; find-in-page (Ctrl+F); GUI terminal copy/paste + font picker ([#​69519](https://github.com/NousResearch/hermes-agent/pull/69519), [#​73121](https://github.com/NousResearch/hermes-agent/pull/73121), [#​72235](https://github.com/NousResearch/hermes-agent/pull/72235), [#​73705](https://github.com/NousResearch/hermes-agent/pull/73705), [#​76395](https://github.com/NousResearch/hermes-agent/pull/76395) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1))
##### Composer & UX
- Attach files/folders/links via picker; composer chips for @​ paths and pasted links; composer undo stack; double-ESC discards draft; double-Enter sends the queued turn; type-to-focus ([#​74668](https://github.com/NousResearch/hermes-agent/pull/74668), [#​73110](https://github.com/NousResearch/hermes-agent/pull/73110), [#​72201](https://github.com/NousResearch/hermes-agent/pull/72201), [#​72288](https://github.com/NousResearch/hermes-agent/pull/72288), [#​74736](https://github.com/NousResearch/hermes-agent/pull/74736), [#​73101](https://github.com/NousResearch/hermes-agent/pull/73101), [#​68918](https://github.com/NousResearch/hermes-agent/pull/68918) — [@​OutThisLife](https://github.com/OutThisLife))
- 2-keypress model switching (⌘⇧M); YOLO in ⌘K with live toggle state; keyboard-first pickers; keyboard navigation for clarify choices; server-owned pins that follow you between apps ([#​74545](https://github.com/NousResearch/hermes-agent/pull/74545), [#​74674](https://github.com/NousResearch/hermes-agent/pull/74674), [#​74602](https://github.com/NousResearch/hermes-agent/pull/74602), [#​69799](https://github.com/NousResearch/hermes-agent/pull/69799), [#​74234](https://github.com/NousResearch/hermes-agent/pull/74234) — [@​OutThisLife](https://github.com/OutThisLife))
- Grouped, live-ticking tool-activity line; improved tool call detail views; [@​session](https://github.com/session) links resolve to clickable titles; brand icons on known-domain links; iMessage-style emoji reactions (opt-in, two-way); double-click to heart ([#​72893](https://github.com/NousResearch/hermes-agent/pull/72893), [#​69868](https://github.com/NousResearch/hermes-agent/pull/69868), [#​71162](https://github.com/NousResearch/hermes-agent/pull/71162), [#​73047](https://github.com/NousResearch/hermes-agent/pull/73047), [#​74533](https://github.com/NousResearch/hermes-agent/pull/74533), [#​74644](https://github.com/NousResearch/hermes-agent/pull/74644) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1))
- Sidebar date dividers + pinned section + opt-in stale-session auto-archive; sessions stop lying about running state; credit-usage toasts; configurable attachment size limit; Cron Blueprints + Webhooks pages; searchable timezone picker ([#​70822](https://github.com/NousResearch/hermes-agent/pull/70822), [#​72303](https://github.com/NousResearch/hermes-agent/pull/72303), [#​69828](https://github.com/NousResearch/hermes-agent/pull/69828), [#​73221](https://github.com/NousResearch/hermes-agent/pull/73221), [#​70066](https://github.com/NousResearch/hermes-agent/pull/70066), [#​69687](https://github.com/NousResearch/hermes-agent/pull/69687), [#​73505](https://github.com/NousResearch/hermes-agent/pull/73505) — [@​OutThisLife](https://github.com/OutThisLife), [@​austinpickett](https://github.com/austinpickett), [@​Adolanium](https://github.com/Adolanium), [@​teknium1](https://github.com/teknium1))
- RFC 8252 native desktop sign-in (system browser + PKCE, no webview cookies); "Connect to existing Hermes" in first-run onboarding; profile-correct pairing approvals with a desktop surface ([#​67920](https://github.com/NousResearch/hermes-agent/pull/67920), [#​70907](https://github.com/NousResearch/hermes-agent/pull/70907), [#​74446](https://github.com/NousResearch/hermes-agent/pull/74446) — [@​benbarclay](https://github.com/benbarclay), [@​OutThisLife](https://github.com/OutThisLife))
- Keep-computer-awake toggle + notch wake indicator; /battery status-bar toggle; UI zoom 90% default preset; status bar hideable ([#​68140](https://github.com/NousResearch/hermes-agent/pull/68140), [#​76396](https://github.com/NousResearch/hermes-agent/pull/76396), [#​68860](https://github.com/NousResearch/hermes-agent/pull/68860), [#​73161](https://github.com/NousResearch/hermes-agent/pull/73161), [#​72960](https://github.com/NousResearch/hermes-agent/pull/72960) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1))
##### Desktop performance (60fps wave 2)
- Streaming cost independent of transcript length; 60fps on real sessions (reflow-gated pins, adaptive flush); drag at 60fps with five streaming tabs; multitab streaming made fast ([#​71835](https://github.com/NousResearch/hermes-agent/pull/71835), [#​72504](https://github.com/NousResearch/hermes-agent/pull/72504), [#​72346](https://github.com/NousResearch/hermes-agent/pull/72346), [#​71780](https://github.com/NousResearch/hermes-agent/pull/71780) — [@​OutThisLife](https://github.com/OutThisLife))
- Hidden-pane timers paused (agents view, cron sidebar, floating pet), scroll/status loops stopped in busy sessions ([#​77651](https://github.com/NousResearch/hermes-agent/pull/77651) — [@​kshitijk4poor](https://github.com/kshitijk4poor)); idle CPU near zero in the background; sidebar/overlay render churn killed; statusbar + transcript stop re-rendering per token/sash-drag/session-switch; ⌘K opens instantly; renderer cold start keeps shiki/mermaid off the boot path ([#​75218](https://github.com/NousResearch/hermes-agent/pull/75218), [#​73698](https://github.com/NousResearch/hermes-agent/pull/73698), [#​72163](https://github.com/NousResearch/hermes-agent/pull/72163), [#​72245](https://github.com/NousResearch/hermes-agent/pull/72245), [#​72524](https://github.com/NousResearch/hermes-agent/pull/72524), [#​74665](https://github.com/NousResearch/hermes-agent/pull/74665), [#​73024](https://github.com/NousResearch/hermes-agent/pull/73024) — [@​OutThisLife](https://github.com/OutThisLife))
- State diagnostics (render + store churn counters) + a lint rule banning atom-mirrored refs so the stale-read bug class cannot return; Playwright E2E suite with visual regression diffs ([#​71925](https://github.com/NousResearch/hermes-agent/pull/71925), [#​71560](https://github.com/NousResearch/hermes-agent/pull/71560), [#​65805](https://github.com/NousResearch/hermes-agent/pull/65805) — [@​OutThisLife](https://github.com/OutThisLife), [@​teknium1](https://github.com/teknium1), [@​ethernet8023](https://github.com/ethernet8023))
##### 🖥️ CLI, TUI & Dashboard
- `!` shell mode; `/init` AGENTS.md generation; `/diff` (staged/all/session, cross-surface); `/context` breakdown; `/focus` reduced-output view; Ctrl+S prompt stash; persistent `/goal` indicator; multi-select clarify (checkboxes) across CLI/gateway/TUI ([#​72257](https://github.com/NousResearch/hermes-agent/pull/72257), [#​72178](https://github.com/NousResearch/hermes-agent/pull/72178), [#​72240](https://github.com/NousResearch/hermes-agent/pull/72240), [#​72242](https://github.com/NousResearch/hermes-agent/pull/72242), [#​72302](https://github.com/NousResearch/hermes-agent/pull/72302), [#​72262](https://github.com/NousResearch/hermes-agent/pull/72262), [#​72244](https://github.com/NousResearch/hermes-agent/pull/72244), [#​72188](https://github.com/NousResearch/hermes-agent/pull/72188) — [@​teknium1](https://github.com/teknium1), salvaging [@​SHL0MS](https://github.com/SHL0MS), [@​iRonin](https://github.com/iRonin), [@​gigi206](https://github.com/gigi206) + more)
- `hermes import-agent` — one-command migration from Claude Code / Codex CLI setups ([#​72190](https://github.com/NousResearch/hermes-agent/pull/72190) — [@​teknium1](https://github.com/teknium1))
- Per-turn summary line + live token flow in the spinner; cross-surface theme SDK (one skin themes CLI, TUI, and desktop, live) ([#​72246](https://github.com/NousResearch/hermes-agent/pull/72246), [#​68857](https://github.com/NousResearch/hermes-agent/pull/68857) — [@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife))
- TUI: reach the model picker without wrecking your draft + mid-turn switching; slash menu leads with your most-used skills; attachments live in the composer; Arabic (ar) locale with RTL across desktop/dashboard/agent ([#​74756](https://github.com/NousResearch/hermes-agent/pull/74756), [#​75931](https://github.com/NousResearch/hermes-agent/pull/75931), [#​75210](https://github.com/NousResearch/hermes-agent/pull/75210), [#​70870](https://github.com/NousResearch/hermes-agent/pull/70870) — [@​OutThisLife](https://github.com/OutThisLife))
- `hermes -w` startup \~14s → \~1.8s; global `--version` fast path; banner update-check 6× faster; dashboard lazy-loads routes + GROUP BY session stats; session filtering tabs (Chats/Automation/All) ([#​71637](https://github.com/NousResearch/hermes-agent/pull/71637), [#​62096](https://github.com/NousResearch/hermes-agent/pull/62096), [#​74188](https://github.com/NousResearch/hermes-agent/pull/74188), [#​72294](https://github.com/NousResearch/hermes-agent/pull/72294), [#​73362](https://github.com/NousResearch/hermes-agent/pull/73362), [#​73865](https://github.com/NousResearch/hermes-agent/pull/73865) — [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Runtime: Node 26 required across installers/heal/upgrade, managed Node/uv resolve before bare PATH, outdated managed trees heal to target major; brew + pip/PyPI wheel channels retired (shell installer / Docker / Nix are the supported channels) ([#​76459](https://github.com/NousResearch/hermes-agent/pull/76459), [#​68217](https://github.com/NousResearch/hermes-agent/pull/68217) — [@​ethernet8023](https://github.com/ethernet8023))
##### 🧩 Skills, Plugins & MCP
- **A2A v1.0** — Agent-to-Agent protocol plugin (closes [#​514](https://github.com/NousResearch/hermes-agent/issues/514)) ([#​77109](https://github.com/NousResearch/hermes-agent/pull/77109) — [@​teknium1](https://github.com/teknium1))
- Curator: surface unmanaged skills + `curator adopt`; skill-description truncation surfaced to authors; grounded-citations skill (+ fact-checking mode); simplify-code v1.1; tldraw-offline scripting skill ([#​71648](https://github.com/NousResearch/hermes-agent/pull/71648), [#​70519](https://github.com/NousResearch/hermes-agent/pull/70519), [#​71698](https://github.com/NousResearch/hermes-agent/pull/71698), [#​77104](https://github.com/NousResearch/hermes-agent/pull/77104), [#​70440](https://github.com/NousResearch/hermes-agent/pull/70440), [#​66896](https://github.com/NousResearch/hermes-agent/pull/66896) — [@​teknium1](https://github.com/teknium1))
- Office skills bundled: docx, xlsx, pdf + refreshed powerpoint; skills-tree debloat continues (yuanbao, segment-anything, jupyter, heartmula, audiocraft → optional-skills; claude-marketplace source removed; hub restructure absorbing themes/desktop-plugins/tui-widgets) ([#​68595](https://github.com/NousResearch/hermes-agent/pull/68595), [#​70452](https://github.com/NousResearch/hermes-agent/pull/70452)–[#​70456](https://github.com/NousResearch/hermes-agent/pull/70456), [#​73903](https://github.com/NousResearch/hermes-agent/pull/73903) — [@​teknium1](https://github.com/teknium1))
- MCP: Comfy Cloud catalog entry with curated 20-tool default; hidden-whitespace warnings in MCP config; pinecone-research optional skill ([#​66112](https://github.com/NousResearch/hermes-agent/pull/66112), [#​75736](https://github.com/NousResearch/hermes-agent/pull/75736), [#​70512](https://github.com/NousResearch/hermes-agent/pull/70512) — [@​teknium1](https://github.com/teknium1))
- MCP lazy server startup from a fingerprint-keyed on-disk tool-schema cache — configured servers no longer all boot at session start (design from [#​56832](https://github.com/NousResearch/hermes-agent/issues/56832)) ([#​77511](https://github.com/NousResearch/hermes-agent/pull/77511) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
- NeMo Relay observability integration — re-landed after an in-window revert, on stable NeMo Relay 0.6 ([#​67607](https://github.com/NousResearch/hermes-agent/pull/67607) — [@​afourniernv](https://github.com/afourniernv))
- Gateway health & diagnostics OTLP export ([#​64536](https://github.com/NousResearch/hermes-agent/pull/64536) — [@​victor-kyriazakos](https://github.com/victor-kyriazakos))
##### 🔒 Security & Reliability
- Iron-proxy credential-injection egress firewall re-landed ([#​70848](https://github.com/NousResearch/hermes-agent/pull/70848) — [@​teknium1](https://github.com/teknium1))
- DNS-pinned SSRF-safe fetches + Slack CDN allowlist; strict redaction at compaction boundaries; ReDoS eliminated in config-key redaction patterns; prose words embedding a secret keyword no longer masked ([#​70193](https://github.com/NousResearch/hermes-agent/pull/70193), [#​69294](https://github.com/NousResearch/hermes-agent/pull/69294), [#​76083](https://github.com/NousResearch/hermes-agent/pull/76083), [#​67776](https://github.com/NousResearch/hermes-agent/pull/67776) — [@​teknium1](https://github.com/teknium1))
- Tier-3 credential reads scoped (FAL/XAI/VERCEL/DAYTONA/GITHUB presence checks etc.); CVE dependency pins refreshed (cryptography, starlette, python-multipart); hindsight env file 0600; /model moved off the gateway event loop ([#​75888](https://github.com/NousResearch/hermes-agent/pull/75888), [#​72362](https://github.com/NousResearch/hermes-agent/pull/72362) — [@​teknium1](https://github.com/teknium1))
- Windows hardening wave: text-mode subprocess decode bug class closed repo-wide, console flashes hidden across daemons/env probes/LSP/installer paths, residual encoding gaps (MCP stdio, gateway update I/O, STT/TTS, desktop spawn) ([#​70875](https://github.com/NousResearch/hermes-agent/pull/70875), [#​70205](https://github.com/NousResearch/hermes-agent/pull/70205), [#​70264](https://github.com/NousResearch/hermes-agent/pull/70264), [#​71014](https://github.com/NousResearch/hermes-agent/pull/71014) — [@​teknium1](https://github.com/teknium1), salvaging several community PRs)
- State/session integrity: four session-state fixes (safe close tracking, flush-cursor class fix, row-retry, usage-PK healer); compact v23 FTS layout + `hermes sessions optimize` + CJK-bigram FTS; read-path split with per-thread read-only connections ([#​75883](https://github.com/NousResearch/hermes-agent/pull/75883), [#​65798](https://github.com/NousResearch/hermes-agent/pull/65798), [#​69423](https://github.com/NousResearch/hermes-agent/pull/69423), [#​73344](https://github.com/NousResearch/hermes-agent/pull/73344) — [@​teknium1](https://github.com/teknium1), [@​kshitijk4poor](https://github.com/kshitijk4poor))
- OpenViking memory-provider hardening — fail closed on blocked endpoints, server verification before credentials are sent, config.yaml-first settings ([#​77747](https://github.com/NousResearch/hermes-agent/pull/77747) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Credential pool: reset-aware primary restore (stay on fallback until the rate-limit window resets) + deferred-refresh locking fixes; FTS UPDATE triggers narrowed with fail-closed CJK migration ([#​77631](https://github.com/NousResearch/hermes-agent/pull/77631), [#​77628](https://github.com/NousResearch/hermes-agent/pull/77628) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
- Config-driven memory allocator trim with telemetry; holographic memory vectors stored float32; loop-invariant HRR encodes hoisted ([#​76905](https://github.com/NousResearch/hermes-agent/pull/76905), [#​76917](https://github.com/NousResearch/hermes-agent/pull/76917), [#​76881](https://github.com/NousResearch/hermes-agent/pull/76881) — [@​kshitijk4poor](https://github.com/kshitijk4poor))
##### 🐛 Notable Bug Fixes
- Voice: full-duplex interruption during generation AND playback; whole-turn desktop speech; auto-TTS delivery gaps ([#​74223](https://github.com/NousResearch/hermes-agent/pull/74223), [#​69936](https://github.com/NousResearch/hermes-agent/pull/69936), [#​73508](https://github.com/NousResearch/hermes-agent/pull/73508) — [@​teknium1](https://github.com/teknium1))
- Desktop: Stop parks the queue instead of firing the next queued prompt; branch-in-new-chat restart loss; false remote-gateway reauthentication; cross-session composer leaks ([#​68725](https://github.com/NousResearch/hermes-agent/pull/68725), [#​71960](https://github.com/NousResearch/hermes-agent/pull/71960), [#​68250](https://github.com/NousResearch/hermes-agent/pull/68250), [#​70986](https://github.com/NousResearch/hermes-agent/pull/70986) — [@​SHL0MS](https://github.com/SHL0MS), [@​alelpoan](https://github.com/alelpoan), [@​helix4u](https://github.com/helix4u), [@​OutThisLife](https://github.com/OutThisLife))
- Gateway: session lists scoped before limiting; relay-backed home delivery after restart; timeline display events persisted ([#​65509](https://github.com/NousResearch/hermes-agent/pull/65509), [#​70102](https://github.com/NousResearch/hermes-agent/pull/70102), [#​69771](https://github.com/NousResearch/hermes-agent/pull/69771) — [@​GodsBoy](https://github.com/GodsBoy), [@​victor-kyriazakos](https://github.com/victor-kyriazakos), [@​ethernet8023](https://github.com/ethernet8023))
- Agent: context-length fallback logging + batch trajectory durability; Codex OAuth context windows revalidated against the live catalog ([#​76027](https://github.com/NousResearch/hermes-agent/pull/76027), [#​68554](https://github.com/NousResearch/hermes-agent/pull/68554) — [@​kshitijk4poor](https://github.com/kshitijk4poor), [@​teknium1](https://github.com/teknium1))
- ...plus roughly 770 more `fix:` PRs across every subsystem this window.
##### 👥 Contributors
**647 contributors** shipped this release (commit authors, co-authors, and salvaged-PR credits).
##### Core
[@​teknium1](https://github.com/teknium1), [@​OutThisLife](https://github.com/OutThisLife) (desktop, voice, perf), [@​kshitijk4poor](https://github.com/kshitijk4poor) (perf, caching, salvage), [@​ethernet8023](https://github.com/ethernet8023) (runtime, E2E, desktop), [@​benbarclay](https://github.com/benbarclay) (relay, HSP, auth)
##### All Contributors (alphabetical)
[@​02356abc](https://github.com/02356abc), [@​0301chris](https://github.com/0301chris), [@​0xAlcibiades](https://github.com/0xAlcibiades), [@​0xDevNinja](https://github.com/0xDevNinja), [@​0xLeathery](https://github.com/0xLeathery), [@​0xprincess](https://github.com/0xprincess), [@​0xr00tf3rr3t](https://github.com/0xr00tf3rr3t), [@​100yenadmin](https://github.com/100yenadmin),
[@​2001Y](https://github.com/2001Y), [@​3ssiri](https://github.com/3ssiri), [@​55nx954gn6-debug](https://github.com/55nx954gn6-debug), [@​686f6c61](https://github.com/686f6c61), [@​87degrees](https://github.com/87degrees), [@​aaronlab](https://github.com/aaronlab), [@​abundantbeing](https://github.com/abundantbeing), [@​Adolanium](https://github.com/Adolanium),
[@​adriansotomora](https://github.com/adriansotomora), [@​adurham](https://github.com/adurham), [@​afourniernv](https://github.com/afourniernv), [@​afurm](https://github.com/afurm), [@​AgenticSpark](https://github.com/AgenticSpark), [@​ahmadashfq](https://github.com/ahmadashfq), [@​AhmetArif0](https://github.com/AhmetArif0), [@​ai-ag2026](https://github.com/ai-ag2026),
[@​AIalliAI](https://github.com/AIalliAI), [@​aider4ryder](https://github.com/aider4ryder), [@​airclear](https://github.com/airclear), [@​ajzrva-sys](https://github.com/ajzrva-sys), [@​akattelu](https://github.com/akattelu), [@​AKAZIK-py](https://github.com/AKAZIK-py), [@​akb4q](https://github.com/akb4q), [@​akshan-main](https://github.com/akshan-main), [@​AlanBurningsuit](https://github.com/AlanBurningsuit),
[@​alelpoan](https://github.com/alelpoan), [@​AlexFucuson9](https://github.com/AlexFucuson9), [@​AlexxRussell](https://github.com/AlexxRussell), [@​AllardQuek](https://github.com/AllardQuek), [@​alt-glitch](https://github.com/alt-glitch), [@​aman-merchant](https://github.com/aman-merchant), [@​amanning3390](https://github.com/amanning3390), [@​amathxbt](https://github.com/amathxbt),
[@​aml1973](https://github.com/aml1973), [@​amoreno16003](https://github.com/amoreno16003), [@​AndrewMoryakov](https://github.com/AndrewMoryakov), [@​andrexibiza](https://github.com/andrexibiza), [@​andynguyendk](https://github.com/andynguyendk), [@​andyylin](https://github.com/andyylin), [@​aneym](https://github.com/aneym), [@​angelos](https://github.com/angelos),
[@​aniruddhaadak80](https://github.com/aniruddhaadak80), [@​AnnasMazhar](https://github.com/AnnasMazhar), [@​annguyenNous](https://github.com/annguyenNous), [@​anoopmehendale-cue](https://github.com/anoopmehendale-cue), [@​AnthonyFrancis](https://github.com/AnthonyFrancis), [@​arcabotai](https://github.com/arcabotai), [@​ArcherQAQ](https://github.com/ArcherQAQ),
[@​Ares4Tech](https://github.com/Ares4Tech), [@​arimu1](https://github.com/arimu1), [@​arnoldfrancisca](https://github.com/arnoldfrancisca), [@​asimons81](https://github.com/asimons81), [@​asorry75](https://github.com/asorry75), [@​AtakanGs](https://github.com/AtakanGs), [@​ATran28](https://github.com/ATran28), [@​austinpickett](https://github.com/austinpickett),
[@​Automata-intelligentsia](https://github.com/Automata-intelligentsia), [@​awain7](https://github.com/awain7), [@​aweiker](https://github.com/aweiker), [@​aydnOktay](https://github.com/aydnOktay), [@​ayushere](https://github.com/ayushere), [@​b](https://github.com/b), [@​baau](https://github.com/baau), [@​baauzi](https://github.com/baauzi), [@​baenregod](https://github.com/baenregod),
[@​bakhtiersizhaev](https://github.com/bakhtiersizhaev), [@​Baophan00](https://github.com/Baophan00), [@​baoyu0](https://github.com/baoyu0), [@​Bartok9](https://github.com/Bartok9), [@​basilalshukaili](https://github.com/basilalshukaili), [@​BB-light](https://github.com/BB-light), [@​bbopen](https://github.com/bbopen), [@​Beandon13](https://github.com/Beandon13),
[@​beardedeagle](https://github.com/beardedeagle), [@​bedirhancode](https://github.com/bedirhancode), [@​benbarclay](https://github.com/benbarclay), [@​benegessarit](https://github.com/benegessarit), [@​benjamin2026-dot](https://github.com/benjamin2026-dot), [@​bennybuoy](https://github.com/bennybuoy), [@​BenSheridanEdwards](https://github.com/BenSheridanEdwards),
[@​BKStock](https://github.com/BKStock), [@​BlackishGreen33](https://github.com/BlackishGreen33), [@​bnikanjam](https://github.com/bnikanjam), [@​bounce12340](https://github.com/bounce12340), [@​Bounty13](https://github.com/Bounty13), [@​bpross](https://github.com/bpross), [@​briandevans](https://github.com/briandevans), [@​bricelb](https://github.com/bricelb), [@​brunopirz](https://github.com/brunopirz),
[@​bryanneva](https://github.com/bryanneva), [@​byshubham](https://github.com/byshubham), [@​camaleonidas](https://github.com/camaleonidas), [@​canorionen](https://github.com/canorionen), [@​carbongotfound](https://github.com/carbongotfound), [@​carljborg](https://github.com/carljborg), [@​carlotestor](https://github.com/carlotestor), [@​carrion256](https://github.com/carrion256),
[@​caseyanthony](https://github.com/caseyanthony), [@​cat-thats-fat](https://github.com/cat-thats-fat), [@​Cdddo](https://github.com/Cdddo), [@​ceverson70](https://github.com/ceverson70), [@​chancelu](https://github.com/chancelu), [@​chaos-xxl](https://github.com/chaos-xxl), [@​CharlesMcquade](https://github.com/CharlesMcquade), [@​chazmaniandinkle](https://github.com/chazmaniandinkle),
[@​chefboyrdave21](https://github.com/chefboyrdave21), [@​chelsealong](https://github.com/chelsealong), [@​Christopher-Schulze](https://github.com/Chr…
randlee
pushed a commit
to randlee/hermes-agent
that referenced
this pull request
Aug 11, 2026
…ames, reply_to context, hello command manifest (NousResearch#71624) - RelayAdapter.create_handoff_thread → one op-gated thread_create op (Discord channel thread / Telegram forum topic / Slack named seed root); None fallback contract preserved for the handoff watcher. - RelayAdapter.rename_thread → thread_rename with only_if_current_name no-clobber guard on the wire (connector enforces; Telegram guarded renames fail safe). The native semantic-rename lane (_is_discord_auto_thread_lane) lights over the relay via the connector-stamped auto_thread_created/auto_thread_initial_name markers parsed onto SessionSource in _event_from_wire. - reply_to {text, author, is_own} wire parse onto the SAME MessageEvent reply-context fields native adapters populate. - gateway/relay/command_manifest.py: the gateway-declared slash-command manifest (native Discord tree mirror) sent on the DISCORD hello; the connector reconciles Discord's global registration (additive field, older connectors ignore). - Contract doc §OutboundAction ops + Phase 4 semantics sections. - 12 new tests (tests/gateway/relay/test_relay_threads.py); relay suite 266 passed.
randlee
pushed a commit
to randlee/hermes-agent
that referenced
this pull request
Aug 11, 2026
- New page user-guide/messaging/relay.md: what Relay is, enrollment (hermes gateway enroll), capability handshake (media, native approval/clarify, threads, typing), config keys, troubleshooting — derived from gateway/relay/ + gateway_enroll.py (PRs NousResearch#48147 NousResearch#48242 NousResearch#71300 NousResearch#71363 NousResearch#71404 NousResearch#71624 NousResearch#69721). - desktop.md: artifacts viewer, timeline rail, find-in-page, multi-window, git review/worktrees, multi-terminal + persistence, theme import, rebindable shortcuts, quick entry, context-usage popover, keep-awake, command palette, Hermes Cloud mode, memory graph — all verified against apps/desktop/ source.
33hodl
pushed a commit
to 33hodl/hermes-agent
that referenced
this pull request
Aug 12, 2026
- New page user-guide/messaging/relay.md: what Relay is, enrollment (hermes gateway enroll), capability handshake (media, native approval/clarify, threads, typing), config keys, troubleshooting — derived from gateway/relay/ + gateway_enroll.py (PRs NousResearch#48147 NousResearch#48242 NousResearch#71300 NousResearch#71363 NousResearch#71404 NousResearch#71624 NousResearch#69721). - desktop.md: artifacts viewer, timeline rail, find-in-page, multi-window, git review/worktrees, multi-terminal + persistence, theme import, rebindable shortcuts, quick entry, context-usage popover, keep-awake, command palette, Hermes Cloud mode, memory graph — all verified against apps/desktop/ source.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Relay Phase 4 — Thread lifecycle & long tail (gateway side)
Gateway half of the Phase 4 pair: NousResearch/gateway-gateway#178. Final phase of the parity plan (Phases 0–3 merged).
RelayAdapter thread ops (op-gated)
create_handoff_thread(parent, name)→ onethread_createop (Discord channel thread / Telegram forum topic / Slack named seed root). Preserves the native contract exactly:Noneon unavailability/decline so the handoff watcher falls back to the parent channel. Acceptsthread_idormessage_idoff the result (Slack seed shape).rename_thread(thread_id, name, only_if_current_name=…, parent_chat_id=…)→thread_renamewith the no-clobber guard on the wire — the connector enforces it platform-side (Discord reads current name; Telegram guarded renames fail safe).Falseon decline, best-effort by contract.Semantic-rename lane over the relay
_event_from_wirenow parses the connector-stampedauto_thread_created/auto_thread_initial_nameontoSessionSource— the native_is_discord_auto_thread_lanecheck lights unchanged (same field contract, no relay special-case), so LLM session titles rename relay auto-threads with the same only-if-current-name guard as native.reply_tocontext parseWire
reply_to {text, author, is_own}maps onto the same MessageEvent fields native adapters populate (reply_to_text/reply_to_author_name/reply_to_is_own_message) — run.py's reply-context injection works identically over the relay. Absent/partial fields degrade to the current behavior.Gateway-declared Discord command manifest
gateway/relay/command_manifest.py— the relay lane's slash-command declaration, mirroring the native Discord tree (same names/descriptions, 27 commands incl./approve,/deny,/steer,/queue). Sent on the Discord hello only; the connector (which holds the bot token) syncs Discord's global registration. Additive field — older connectors ignore it. Interactions still dispatch through the existing passthrough plane; no new handler surface.Contract doc
docs/relay-connector-contract.md: op table rows forthread_create/thread_rename+ Phase 4 semantics sections (thread lifecycle, auto-thread markers, command manifest, reply_to enrichment) + non-goals documented (Discord voice, Slack App Home, Telegram draft-API streaming, Baileys personal-account semantics — native-only by design).Tests
12 new (
tests/gateway/relay/test_relay_threads.py): thread op routing + op-gating + decline fallbacks, guard on the wire, marker lane parity, reply_to parse (full/partial/absent), manifest naming rules + native-mirror spot-check, Discord-only hello manifest. Relay suite: 266 passed.Deploy notes