- context_engine.py: document that select_context() runs before cache-control
and all request sanitizers, so (a) replacements still pass host validation
and (b) the no-op default keeps the request byte-stable (AGENTS.md prompt-
cache invariant). Note the hook is evaluated per provider request.
- tests: no-op path is byte-stable for cache-control; a role-unusual
replacement is passed through for the existing downstream sanitizers to
normalize (select_context does structural validation only).
The on_turn_complete() observation hook is the engine's post-turn signal,
so it should receive the completed turn's canonical token usage when the
host has it, not a hardcoded None. Per @johnnykor82's #41918 contract: the
engine uses prompt/completion + cache_read/write/reasoning buckets to judge
how large/expensive the selected context was before the next select_context().
- conversation_loop.py: stash the most recent provider response's usage_dict
(the same canonical shape fed to update_from_response) on the agent as
_last_turn_usage; reset to None at turn start so turns that never reach a
provider response (early failure / interrupt) forward None, not a stale
prior turn's usage.
- turn_finalizer.py: forward agent._last_turn_usage instead of usage=None.
- context_engine.py: document the usage param contract on the ABC hook.
- tests: cover both ends through the real finalize_turn path — completed turn
forwards the full canonical bucket set intact; no-response turn forwards None.
Co-authored-by: johnnykor82 <johnnykor82@users.noreply.github.com>
Adds the post-turn observation verb as the companion to select_context():
an optional, no-op-default on_turn_complete() called once after the
assistant/tool loop finishes, with the finalized transcript snapshot. Lets
an engine ingest/index/summarize the completed turn to inform the next
select_context(). Wired via _notify_context_engine_turn_complete() from
turn_finalizer.finalize_turn(); fail-open, base no-op short-circuited so
non-implementing engines (incl. the built-in compressor) pay nothing.
This is the request-assembly + observation pair from #41918; with this
commit the PR fully subsumes #41918's two hooks (prepare_request_messages
-> select_context, on_turn_complete) rather than only the selection half.
Co-authored-by: johnnykor82 <johnnykor82@users.noreply.github.com>
Adds an optional, no-op-default select_context() hook to the ContextEngine
ABC, called every turn after the request messages are assembled and before
provider dispatch — independent of should_compress(). Lets an engine select
or replace which context enters the prompt for a single request (retrieval,
topic routing, role/branch switching) without mutating persisted history,
removing the need to abuse should_compress()=True as a per-turn callback.
The host call site (_apply_context_engine_selection) is fail-open: a missing
hook, an exception, or an invalid return value leaves the assembled request
untouched. Additive and non-breaking: the built-in compressor and every
existing engine are unaffected.
Consolidates the per-turn request-assembly surface proposed across #41918,
Related: #36765#41918#24949#47109#50053#23837#25115#29370
Routine automatic compression stays silent-by-design on chat platforms
(default unchanged, byte-identical). New opt-in config key
compression.progress_notices (bool, default false) opens a gate on the
gateway noise filter (_prepare_gateway_status_message) that lets ROUTINE
compression progress statuses through to chat surfaces.
- Membership is derived from the #69550 status template constants in
agent/conversation_compression.py (compiled to a literal-escaped regex,
never re-inlined wording), so unrelated noisy statuses (aux failures,
provider retry/rate-limit chatter) stay suppressed even when enabled.
- The compaction completion notice (COMPACTION_DONE_STATUS, #69546
lifecycle 'compacted' edge) already flows through the status path and
passes the filter — no new emit site needed.
- Config wired everywhere: hermes_cli/config.py DEFAULT_CONFIG,
cli-config.yaml.example, gateway raw-YAML read (live, mtime-cached),
gateway hot-reload cache-busting key list, website configuration docs.
- Failure notices and manual /compress feedback remain always-visible;
VISIBLE_COMPRESSION_MESSAGES and emit sites untouched.
Design by @havok-training (issue #52995).
_render_tool_calls only handled dict-shaped entries; a SimpleNamespace-
shaped tool_call (SDK-style stream-stitched responses) rendered as
'[called tool: tool]', silently losing the function name and arguments
from the advisory view. Handle both shapes (including a namespace-shaped
nested function inside a dict entry).
One-hunk hardening salvaged from closed#59712.
Co-authored-by: SquabbyZ <601709253@qq.com>
Reference models may have a smaller context window than the aggregator
(e.g. kimi-k2.7-code @ 262K advising a glm-5.2 @ 1M conversation).
Without context-length protection, a reference whose window is exceeded
gets a hard HTTP 400 from the provider, which _run_reference's
try/except silently converts to a [failed: …] note — the MoA turn
silently degrades to fewer references (#60345).
Redesigned implementation of #60387:
- Estimate AFTER the advisory system prompt is prepended, so the
request that is actually sent is what gets budgeted.
- Reserve output headroom: the preset's reference_max_tokens when set,
else an 8192-token constant, plus a 10% estimator-error fraction.
- Trim on advisory-view boundaries (text-only user/assistant turns; no
tool-result frames to orphan), preserving the system prompt, the
user-first invariant after every pop (never assistant-first), and the
trailing synthetic user turn.
- Cache get_model_context_length per (provider, model) in a per-fan-out
dict shared across the worker threads, so a turn resolves each
window once instead of probing metadata sources
per-reference-per-iteration (failures are cached too).
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Follow-ups for salvaged #56344:
- A reference that completes between the interrupt check and the reap
keeps its REAL output and accounting (the provider call billed) instead
of being zeroed with a placeholder.
- A reference still in flight at interrupt time gets a placeholder in the
results, but its future now carries a done-callback that folds the
eventual real usage/cost into the facade's pending accounting
(late_accounting_sink -> _record_late_reference_accounting), so billed
spend is never silently dropped. Pending totals are folded (not
overwritten) and guarded by a lock since done-callbacks fire on
executor worker threads.
- Interrupted placeholder results are no longer written into the facade's
turn-scoped reference cache: a cache HIT never re-runs references, so
caching a partial snapshot would replay '[skipped: interrupted by
user]' notes for the rest of the turn. The cache is left empty and the
next create() re-runs the fan-out.
agent/tool_executor.py's concurrent tool batch checks agent._interrupt_requested
and aborts the wait early; agent/moa_loop.py's _run_references_parallel had
no equivalent, so a MoA-enabled turn blocked on ThreadPoolExecutor.result()
until every reference model finished or hit its own individual
auxiliary.moa_reference timeout -- there was no way for the user to abort a
live turn mid-fanout.
Thread an optional `agent` parameter through aggregate_moa_context ->
_run_references_parallel (used when MoA references run alongside the main
model) and MoAClient/MoAChatCompletions (used when the MoA preset itself is
the acting model), then poll concurrent.futures.wait() in
_REFERENCE_POLL_INTERVAL_S slices instead of blocking on future.result() per
reference, checking agent._interrupt_requested each cycle.
Deliberately scoped to interrupt/cancel only -- no new or changed timeout
value, so this doesn't overlap open PRs #53784/#53875 (which lower the
per-reference timeout default but don't add interrupt support). `agent` is
optional and defaults to None, so any caller that doesn't pass it keeps
today's uninterruptible blocking behavior unchanged.
Extends the all-references-failed short-circuit (#56975) to the
persistent `provider: moa` facade path: MoAChatCompletions.create()
previously attached 'use the reference responses below' guidance built
entirely from failure sentinels and called the aggregator with it. Now
an all-failed turn attaches either the sanitized unavailability notice
(loud policy) or nothing (silent policy), and the aggregator — which IS
the acting model — simply acts alone. Advisor accounting for the failed
fan-out is still recorded.
Co-authored-by: liuhao1024 <sunsky.lau@gmail.com>
When every MoA reference model returns a failure (HTTP error, timeout,
etc.) or is skipped by the recursion guard, the one-shot aggregator
synthesis call is now skipped entirely. Previously it would try to
synthesise a wall of failure sentinels, which could block for the full
provider timeout (observed ~6 min on SenseNova) before returning a
non-retryable error that left the session hanging.
The early return carries the sanitized unavailability notice (never raw
provider error text, per the failed-reference containment) so the main
agent loop can still act in single-model mode.
Salvaged from #56975, reworked atop the _is_failed_reference helpers.
Follow-ups for salvaged #53784:
- reference_timeout now defaults to None = no per-preset override, so the
reference fan-out inherits auxiliary.moa_reference.timeout (900s default)
via call_llm's own per-task timeout resolution. The PR's 30.0s default
would have cut off long-thinking advisors mid-response, and its 300s max
cap capped legitimate explicit values — both removed. Explicit per-preset
values are still honored as-is.
- _is_failed_reference also treats '[skipped: …]' recursion-guard notes as
internal sentinels, keeping them out of both aggregator prompts.
- Dashboard/desktop TS types updated to number | null; web_server validator
accepts null/empty as 'inherit'.
Keep MoA reference display events off the machine-readable -Q stdout
surface (platform=cli with tool_progress_mode=off) while preserving them
everywhere else. Extracts the relay into module-level helpers so the
policy is testable.
Salvaged from #67334.
Per the 'when in doubt, optional' rule — the live-kernel workflow needs
uv + JupyterLab + a running server + a cloned hamelnb repo, a niche
setup that shouldn't ship active by default.
Renamed jupyter-live-kernel -> jupyter-notebook (skill name, page slugs,
catalogs, sidebar, zh-Hans mirror, darwinian-evolver related_skills).
Install via: hermes skills install official/data-science/jupyter-notebook
Follow-up to the #67690 salvage (@m4r13y). The PR's tools/env_probe.py
hunk was written against the old capture_output=True _run(); #67964/#67999
rewrote _run to temp-file capture on July 20, so that hunk no longer
applied — but the rewritten _run still lacked creationflags and kept
flashing one console per probe (~5 per kanban worker start) from
windowless parents. Re-implement the one-line fix against the current
shape: creationflags=windows_hide_flags() on the temp-file subprocess.run,
preserving the #67964 grandchild-can't-wedge-the-pipe contract.
Also add the tests the PR didn't ship, in
tests/test_windows_subprocess_no_window_flags.py:
- env_probe._run passes CREATE_NO_WINDOW and keeps temp-file (non-PIPE)
stdout/stderr + DEVNULL stdin
- lazy_deps uv install / pip --version probe / pip install fallback /
ensurepip bootstrap all pass CREATE_NO_WINDOW
- suppress_platform_ver_console: POSIX no-op (platform._syscmd_ver
untouched, win32_ver() still returns), and simulated-Windows stubbing
(echo stub installed, idempotent, never raises)
From windowless processes (the pythonw gateway and the kanban workers it
spawns), three spawn paths flash visible console windows on Windows:
1. tools/env_probe.py::_run() ran its interpreter/pip probes
(python3 / python / pip / 'python3 -m pip' / PEP-668 check, ~5 per
worker start) without creationflags — one console flash per probe.
2. tools/lazy_deps.py had four spawn sites with the same defect:
'uv pip install', the 'pip --version' probe, ensurepip, and the
pip install fallback.
Both now pass creationflags=windows_hide_flags() (CREATE_NO_WINDOW on
Windows, 0 on POSIX) — stdio capture still works because the child is
hidden, not detached.
3. CPython 3.11's platform.win32_ver() unconditionally calls
_syscmd_ver(), which runs 'cmd /c ver' via
subprocess.check_output(shell=True) with no window suppression. Any
dependency touching platform.uname()/version()/platform() at import
time flashes one 'cmd' window per windowless process. New helper
_subprocess_compat.suppress_platform_ver_console() (Windows-only,
never raises) stubs platform._syscmd_ver so win32_ver() falls back to
sys.getwindowsversion().platform_version — verified byte-identical
platform.platform() output on CPython 3.11
('Windows-10-10.0.26100-SP0' either way). Called at the top of
hermes_cli/main.py, right after the hermes_bootstrap guard, before
heavyweight imports.
Verified on Windows 11 by polling EnumWindows at ~15 ms and attributing
new visible HWNDs to the suspect process tree (conhost child presence is
NOT evidence of a visible window — it appears even with
CREATE_NO_WINDOW). Tests: tests/tools/test_windows_native_support.py,
test_env_probe.py, test_lazy_deps.py, test_lazy_deps_durable_target.py —
153 passed; the 3 failures are pre-existing on upstream/main in a
Windows environment (POSIX-only assertions and NTFS chmod semantics).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The per-advisor enabled toggle adds enabled=True to normalized slots;
the JSON-string-parse and per-slot max_tokens tests from sibling
clusters asserted exact dicts. Compare against the enabled-augmented
expectation instead.
The per-reference-model enabled toggle (#59753 salvage) intentionally adds
'enabled' to normalized slot dicts. The two endpoint tests asserted the
exact key set {provider, model} — convert them to subset + round-trip
contracts so optional slot keys (enabled, reasoning_effort, max_tokens)
don't break them again.
Follow-up for salvaged PR #59753 rebased over the per-slot
reasoning_effort feature: _clean_slot now round-trips reasoning_effort
AND enabled together; add a normalize→normalize regression test, update
the validate/normalize agreement contract for the canonical enabled
default, restore the desktop per-slot toggle test on the current
autosave editor, and map oppenheimor's contributor email.
Frontend consumers for the events added by PR #59646: the TUI shows a
replace-in-place 'MoA: refs k/n' activity line (swapped for 'MoA:
aggregating…' on the aggregator phase), and desktop streams '◇ MoA refs
k/n' lines into the reasoning disclosure, self-cleaned by the first
moa.reference block.
Adds per-reference progress events and a phase-transition marker to the
MoA display pipeline so TUI / CLI / desktop surfaces can render a status
bar like `MOA: 2/3 refs done` and surface which phase (reference vs
aggregator) is currently active.
- `moa.progress` — fired once per reference completion with
`refs_done`, `refs_total`, and the source label
- `moa.phase` — fired on phase transitions (currently the single
`phase="aggregator"` transition once the fan-out
finishes)
Plumbed through the existing `reference_callback` →
`tool_progress_callback` → gateway path; no new UI surface. The legacy
`moa.reference` / `moa.aggregating` events are unchanged for backwards
compatibility.
AI-assisted fix by https://github.com/SquabbyZ/peaks-loop
Salvaged from PR #59743. Original author email was malformed
(sr@samirusani, not resolvable to a GitHub account), so the commit is
re-authored with credit via trailer.
Co-authored-by: Sami Rusani <samrusani@users.noreply.github.com>
Maintainer review (hermes-sweeper) on this PR found the fix was
incomplete: two paths still hid the MoA reference panel under
thinking: hidden.
1. thinking.tsx: the mount useState correctly seeds openThinking from
(visible.thinking === 'expanded' || reasoningAlwaysVisible), but the
re-sync effect on [visible] fires after the FIRST render too, not
just later updates, and lacks the reasoningAlwaysVisible OR — so it
immediately collapsed a just-opened MoA panel right after mount.
Skip only the effect's very first run (a ref flag); every later
visible change still re-syncs without the override, preserving the
documented no-OR-at-effect-time contract (manual collapse sticks).
2. useMainApp.ts: showProgressArea's streamSegments predicate gated
thinking content on thinkingPanelVisible alone, so an MoA reference
segment (segment.isMoaReference, same flag messageLine.tsx's
shouldShowThinkingTrail already honors per #64657) never kept the
live progress area up when thinking was hidden — StreamingAssistant
then returned early before MessageLine was ever reached. Added the
same override.
Added tests/thinkingMoaReferenceVisibility.test.tsx: mounts ToolTrail
with reasoningAlwaysVisible + sections.thinking: hidden, awaits queued
effects, and asserts the chevron is still open (▾, not ▸) once they
settle.
Validation:
npx vitest run src/__tests__/thinkingMoaReferenceVisibility.test.tsx
-> 1 passed
Fail-before: reverting only the thinking.tsx ref-guard reproduces the
exact regression -- the same test's frame capture shows the panel
open on first paint then collapsing to ▸ once the effect fires, and
the 'not.toContain(▸)' assertion fails as expected.
npx vitest run (full ui-tui suite): 1115 passed, 8 failed -- all 8
pre-existing and unrelated (terminalSetup/terminalParity/editor
resolution env-path tests), confirmed by running them in isolation
with the same result regardless of this diff.
npx tsc --noEmit: clean.
npx eslint src/components/thinking.tsx src/app/useMainApp.ts: clean.
Every moa.reference gateway event stores its labelled reference-model
output in a Msg's generic `thinking` field (turnController's
recordMoaReference), which messageLine.tsx and the ToolTrail component gate
on `display.sections.thinking`'s resolved mode. When that mode resolves to
`hidden`, MoA reference blocks were suppressed along with ordinary model
reasoning — even though (per #53855) references are the mixture-of-agents
process the user explicitly opted into, not private reasoning, and should
stay visible regardless of the thinking-section setting.
Adds Msg.isMoaReference (set by recordMoaReference), a shouldShowThinkingTrail
helper mirroring the existing shouldShowResponseSeparator pattern, and a
reasoningAlwaysVisible prop threaded into ToolTrail to bypass the two
suppression gates (the trail-wrapper return-null check and the
allHidden/panel-push checks) plus the panel's initial open state and the
shift-click expand-all gesture, so a MoA reference panel is not just present
in the tree but actually visible and openable on first paint.
Fixes#64657
Every moa.reference event called appendReasoningDelta(..., replace=true),
which wipes ALL existing reasoning-type message parts and seeds exactly one
new part. With two or more MoA reference models, each later reference
erased the reasoning disclosure built by earlier references, so only the
last advisor's output ever stayed visible instead of one labelled block per
reference (contradicting the multi-reference visibility behavior from
#53855).
Only the first reference (index <= 1, or missing) now replaces — preserving
the original "clear stale reasoning from before this turn" behavior. Every
later reference accumulates via the existing queue-then-flush path instead,
applied immediately since each reference arrives as one complete block
rather than incremental tokens.
Fixes#64658
The extracted findGitBash builds Windows-style candidate paths, but the
vitest suite (and any POSIX CI host) runs with posix path.join, which
mangles 'C:\Program Files' + segments into slash-joined paths and broke
the invalid-override fallback test. Use path.win32.join explicitly so
candidate construction is host-independent.
Port the HERMES_GIT_BASH_PATH env var check from main.cjs to main.ts
after the TS conversion. Also extract findGitBash to a dedicated module
for testability and add focused regression tests for override precedence
and invalid-override fallback.
Regression tests for the #47971 salvage: the LSP language-server spawn
must pass windows_hide_flags() creationflags while keeping PIPE stdio
and start_new_session, and the npm/go LSP auto-installer subprocess.run
calls must carry the same hide flags with DEVNULL stdin and
capture_output intact.
Salvaged from PR #47971 (LSP subset). On Windows, .cmd-wrapped language
servers (e.g. pyright-langserver.CMD launched via cmd.exe /c) and the
npm/go/pip LSP auto-installers spawn without CREATE_NO_WINDOW, so a
console window flashes whenever the spawn happens under a console-less
parent — e.g. a VS Code/Zed extension host running the ACP adapter.
- agent/lsp/client.py::_spawn: pass creationflags=windows_hide_flags()
to the language-server asyncio subprocess (inert 0 on POSIX;
start_new_session is kept — it is POSIX-only and ignored on Windows).
- agent/lsp/install.py: same flags on the npm and go installer
subprocess.run calls. The pip path goes through
hermes_cli.tools_config._pip_install, which already hides its windows.
Adapted from the PR's hand-rolled _NO_WINDOW constant to the repo's
hermes_cli._subprocess_compat.windows_hide_flags() convention.
The Windows /restart watcher's outer Popen spawns the watcher with
windows_detach_popen_kwargs() (which carries CREATE_BREAKAWAY_FROM_JOB),
but a restrictive parent job object can reject that bit with OSError and
the current call has no retry. Preserve the current watcher
implementation and add a focused breakaway-denied fallback.
Preserved from current main: watcher_python / pythonw.exe selection, the
str(restart_after_s) deadline, the scrubbed watcher_env, the intentional
no-breakaway inline respawn, and the entire POSIX setsid/bash path.
- primary keeps **windows_detach_popen_kwargs()
- on OSError, retry the same argv/env with
creationflags=windows_detach_flags_without_breakaway()
- on dual failure, log a definitive, path-safe warning (interpreter
basename + numeric winerror/errno only) and return without crashing
Replace the superseded breakaway-first inline design and its AST tests
with focused behavioral coverage that drives the real coroutine with a
mocked subprocess.Popen (retry, argv/env/DEVNULL preservation, POSIX
single-session kwarg, no-breakaway inline respawn, secret-safe logging).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds moa.privacy_filter ('' | display | full, default off — issue #59959):
- display: redact user-visible surfaces only (reference blocks emitted to
the UI + saved MoA trace records, including per-advisor full input/output
and the aggregator-input copy); the aggregator sees raw advisor text so
synthesis quality is unaffected.
- full: additionally redact the advisor text injected into the aggregator
prompt, on both the persistent facade path and the one-shot /moa
synthesis path (the issue's literal ask). Legacy boolean true maps here.
Secret/credential shapes (API-key prefixes, JWTs, private keys, DB
connection strings) are delegated to the central redactor
(agent.redact.redact_sensitive_text, force=True + code_file=True); the MoA
filter adds only email and clearly delimited phone-number patterns. No
bare 10-digit matching: line numbers, timestamps, epoch values, git SHAs,
IPs, versions, and source-code assignments in code-review-shaped advisory
text pass through byte-identical. The reference cache always holds raw
text — redaction happens at each consuming surface, so a mid-session mode
change never leaks or double-redacts.
Reworked from PR #60463: replaced its hand-rolled pattern list (which
matched bare digit runs and re-implemented key shapes) with central-
redactor reuse + safe patterns, and split the single boolean into
display/full modes. Credited for the feature framing.
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Extends the fanout enum with 'every_n:<N>' (N >= 2): advisors run on the
first iteration of each user turn and every Nth tool iteration after it;
off-cadence iterations REUSE the cached guidance from the last on-cadence
run via the same cache mechanism the user_turn fanout uses, so the
aggregator still gets advice on every step. The cadence counter is scoped
per user turn (resets on a new user message) and only advances when the
advisory state actually changes, so streaming retries never consume a
cadence slot. Mapping form {mode: every_n, n: N} normalizes to the
canonical string. Unknown/degenerate values fall back to per_iteration.
Addresses issue #63393 (advisor fan-out multiplies turn latency/cost by
the tool-iteration count). Redesigned from PR #63448: the submitted shape
skipped references entirely on off-cadence iterations (aggregator ran
advice-less); this version keeps the last advice in play, credited for
the idea and cadence framing.
Config-gated, default-off (default fanout remains per_iteration).
Co-authored-by: webtecnica <75556242+webtecnica@users.noreply.github.com>
Adds test_moa_gemini_aggregator_sanitize_uses_real_model: drives a full MoA
tool-call turn (virtual-provider mode) with a Gemini aggregator and asserts
the strict-API sanitize pass is invoked with the resolved aggregator model
(gemini-3-pro-preview), never the virtual preset name once a slot is
resolved — the exact path that stripped extra_content/thought_signature and
made Gemini aggregators 400 (#65092).
Writing the test surfaced a gap in the salvaged #66212 fix: in virtual-
provider MoA mode (provider=moa, no moa_config threaded through
run_conversation) the conversation-loop branch never fired because it only
consulted moa_config. Extend it to fall back to the facade's
last_aggregator_slot — the same source the handle_max_iterations fix uses —
so both MoA entry modes resolve the real aggregator model.
Also adds the contributors/emails mapping for the #15676 credit base.
Follow-ups to the salvaged core of #60293:
- Gate the x-initiator header on _normalize_aux_provider() instead of a
literal 'copilot' string compare, so slot configs spelled github /
github-copilot / github-models / copilot-acp / mixed case all get the
user-turn attribution.
- Thread extra_headers through _retry_same_provider_sync/_async so the
credential-refresh and pool-rotation retry rebuilds don't silently drop
the header (the rebuilt kwargs previously started from scratch).
- Add a transport-boundary test asserting the header reaches the SDK
client's create() kwargs (no call_llm mocking), an alias-spelling
matrix test, and a retry-rebuild preservation test.
Follow-up to the salvaged core of #53802: a naive MoAClient(preset) rebuild
restores a working facade but silently drops the reference_callback relay
wired in agent_init, so moa.reference / moa.aggregating display events stop
reaching every frontend for the rest of the session.
Introduce agent.moa_loop.build_moa_facade(agent, preset) as the single
construction point for the MoA facade and use it at:
- initial client construction (agent_init.py)
- turn-start fallback restore (restore_primary_runtime)
- transient transport recovery (try_recover_primary_transport — previously
fell through to _create_openai_client with MoA's empty client_kwargs and
died with 'api_key client option must be set')
- mid-session model switches (switch_model)
The relay reads agent.tool_progress_callback at emit time, so callbacks
attached after construction are picked up automatically.
Adds test_moa_restored_facade_still_emits_reference_events covering event
delivery through a restored facade.
When Hermes fails over from a non-Gemini provider (xAI, Anthropic, etc.) to
Gemini mid-conversation, the existing assistant tool_calls in history carry
no Gemini ``extra_content.google.thought_signature`` (the originating provider
never emits one). The native adapter's ``_translate_tool_call_to_gemini``
omitted ``thoughtSignature`` entirely in that case, so Gemini 3 thinking
models rejected every replayed turn with::
HTTP 400 INVALID_ARGUMENT
Function call is missing a thought_signature in functionCall parts.
Additional data, function call default_api:<tool_name>, position N.
The Cloud Code Assist sibling adapter already handles this exact case by
emitting a sentinel ``"skip_thought_signature_validator"`` (see
``agent/gemini_cloudcode_adapter.py:106``, originally added in #11270 and
documented as matching ``opencode-gemini-auth``'s approach). This change
mirrors that fallback in the native adapter so the two paths behave
identically when replaying cross-provider history.
Verified live against ``generativelanguage.googleapis.com/v1beta`` with
``gemini-3-pro-preview``: synthetic 2-turn conversation with no real
``thoughtSignature`` returns 400 without the sentinel and 200 with it.
Test added: ``test_build_native_request_emits_sentinel_for_cross_provider_tool_call``.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When MoA mode is active with a Gemini model as the aggregator,
agent.model holds the virtual preset name (e.g. "closed"), not the
actual aggregator model name. The _sanitize_tool_calls_for_strict_api
call uses agent.model to decide whether to keep extra_content
(thought_signature) on tool_calls — since "closed" doesn't contain
"gemini", the thought_signature is stripped and the Gemini aggregator
rejects the next request with HTTP 400 (INVALID_ARGUMENT):
"Function call is missing a thought_signature in functionCall parts."
Fix: resolve the actual aggregator model name from moa_config
(conversation_loop) or last_aggregator_slot (chat_completion_helpers)
and pass it to _sanitize_tool_calls_for_strict_api so the
_model_consumes_thought_signature check sees the real Gemini model.
Closes#65092
Bare ContextCompressor.__new__ doubles (test_compress_focus,
cross_session_guard, image_tokens, pre_compress_memory_context) skip
__init__ and lack the attribute — the documented compression-path
test-double pitfall. Guard with getattr default 1 + int type pin
(bool excluded).
