Commit Graph

14549 Commits

Author SHA1 Message Date
Teknium eb8a30cc2f simplify(compat): skills_sync/skill_manager/kanban/computer_use — drop 37 re-exports/aliases, repoint 8 callers + 9 test files 2026-09-03 13:27:01 -07:00
Teknium 8e1a5b9f46 simplify(compat): tools/wake_word — drop 5 re-exports, repoint 1 test 2026-09-03 13:24:36 -07:00
Teknium 98c140bc4b simplify(compat): code_execution_tool/environments.local — drop 36 re-exports, repoint 5 callers + 14 test files 2026-09-03 13:24:04 -07:00
Teknium 5c3db6750f simplify(compat): tools/computer_use/cua_backend — drop 24 re-exports + shutil re-import, repoint 5 callers + 14 test files 2026-09-03 13:22:51 -07:00
Teknium 2dcd97d648 simplify(compat): tools/skills_hub — drop 51 re-exports/aliases, repoint 25 callers + 11 test files 2026-09-03 13:22:48 -07:00
Teknium fc9ed6973e simplify(compat): tts_tool — drop 57 re-exports/aliases, repoint 10 callers + 23 test files 2026-09-03 13:20:22 -07:00
Teknium 62732c7d8b simplify(compat): tools/transcription_tools — drop 47 re-exports/aliases, repoint 4 callers + 9 test files 2026-09-03 13:17:25 -07:00
Teknium d179f28307 simplify(compat): anthropic_adapter — drop 30 re-exports + 1 alias, repoint 22 caller files (32 sites), 38 test files (~125 sites) 2026-09-03 13:16:47 -07:00
Teknium 2031c819fe simplify(compat): gateway — re-land 92d0bd0d73 (reverted by stale-index commit b818085298)
Re-applies the gateway compat removal byte-for-byte; see 92d0bd0d73 for the
full inventory (30 re-exports/aliases + 2 shim modules dropped, 3 shim-only
names re-removed, 24 callers + 34 test files repointed). No new changes.
2026-09-03 13:12:50 -07:00
Teknium 89fbd5d4d3 simplify(compat): conversation_loop — drop 9 re-exports, repoint 7 callers, 35 test sites 2026-09-03 13:12:23 -07:00
Teknium b818085298 simplify(compat): doctor/status — drop 13 re-exports + the doctor_* globals() facade (97 names), repoint 6 callers / 13 tests 2026-09-03 13:10:52 -07:00
Teknium 92d0bd0d73 simplify(compat): gateway — drop 30 re-exports/aliases + 2 shim modules, re-remove 3 shim-only names, repoint 24 callers + 34 test files
Per COMPAT_REMOVAL.md (internal import paths are not a stable API):

Re-exports removed
- gateway/run.py: atomic_json_write, load_dotenv, resolve_delivery_transport,
  TurnRunner, merge_pending_message_event, _arm_loop_floor_timer,
  start_loop_liveness_watchdog, DEFAULT_GATEWAY_POST_INTERRUPT_GRACE_TIMEOUT,
  _UNSET (9) — run_* mixins and tests now import from the defining module
  (gateway.delivery / gateway.run_turn_runner / gateway.platforms.base /
  gateway.shutdown_watchdog / gateway.restart / utils).
- gateway/session.py: SessionResetPolicy, normalize_whatsapp_identifier,
  TranscriptReadError, auto_continue_freshness_window (+ "_now & co." noqa
  facade) — gateway/__init__ takes SessionResetPolicy from .config; callers
  take TranscriptReadError from gateway.session_transcript.
- gateway/kanban_watchers.py: _wake_scope_id + "tests import via origin" noqa
  facade; tests import from kanban_watchers_common / _notifier.
- gateway/slash_commands.py: _model_switch_skew_guard, HISTORY_UNREADABLE.
- gateway/stream_consumer.py: escape_code_fences_for_display.
- gateway/platforms/api_server.py: "re-exported" RunIdempotencyStore comment;
  tui_gateway + tests import gateway.platforms.api_server_run_idempotency.
- gateway/platforms/__init__.py: PEP 562 __getattr__/__dir__ lazy QQAdapter /
  YuanbaoAdapter facade (no in-tree importer).
- gateway/startup_watchdog.py: whole re-export shim module deleted; the three
  in-tree callers import hermes_startup_watchdog directly.

Aliases removed
- gateway/platforms/signal.py: SignalAdapter._markdown_to_signal.
- gateway/shutdown_forensics.py: _parse_systemd_duration_to_us.
- gateway/platforms/yuanbao.py: OutboundManager.start_slow_notifier /
  cancel_slow_notifier / get_chat_lock / _chat_locks / CHAT_DICT_MAX_SIZE
  delegates; module-level get_active_adapter / send_yuanbao_direct;
  MarkdownProcessor has_unclosed_fence / ends_with_table_row /
  split_at_paragraph_boundary static pass-throughs (chunk_markdown_text stays —
  it carries yuanbao's chunking policy). tools/send_message_senders +
  tools/yuanbao_tools call YuanbaoAdapter.get_active() / sender.send_direct().

Shim-only names re-removed (earlier review-fix round 96c104c903)
- CapabilityDescriptor.from_platform_entry (+ tests/gateway/relay/test_descriptor_from_entry.py)
- SessionTurnLeaseRegistry.__len__ (+ its test)
- is_relay_media_url KEPT: download() uses it, real internal helper.

Tests that pinned a shim (startup_watchdog re-export identity, api_server
RunIdempotencyStore identity, signal wrapper parity) are dropped; tests that
pinned live behavior are repointed at the implementation.

Note: the gateway/run.py hunk of this change was swept into 00a3cfe5c9 by a
concurrent commit on the shared worktree; this commit carries the rest.

Verified in an isolated worktree (HEAD + this change only): ruff clean,
import-smoke of all 33 touched modules under a fresh HERMES_HOME,
605 passed / 1 skipped across the 39 covering test files.
2026-09-03 13:10:49 -07:00
Teknium 67ccfaed39 simplify(compat): image_generation_tool — drop 2 re-export noqa blocks (13 names), repoint 2 callers + 2 tests to image_generation_catalog 2026-09-03 13:10:38 -07:00
Teknium 2ed36eff40 simplify(compat): web_tools — drop 35 re-exports/aliases, repoint 2 callers + 6 test files 2026-09-03 13:08:53 -07:00
Teknium 204c36fd5c simplify(compat): vision_tools — drop 4 re-exports + 2 shim-only sync validators, repoint 1 caller; SSRF tests now pin the live async validator 2026-09-03 13:07:34 -07:00
Teknium eeb7671e69 simplify(compat): hermes_cli small facades — drop 7 re-exports/aliases (+relay_runtime alias module), repoint 12 callers/tests 2026-09-03 13:05:57 -07:00
Teknium b39948106a simplify(compat): acp_adapter — drop 1 re-export, re-remove 5 shim-only names, repoint 4 tests
server: drop ACP_MAX_MODELS_PER_PROVIDER re-export (test -> model_catalog).
Re-removed from 96c104c903: SessionManager.remove_session/cleanup (+ now-orphaned _clear_task_cwd/_delete_persisted)
and their 4 tests; auth.has_provider (tests assert detect_provider() directly);
edit_approval.clear_/get_edit_approval_requester (test teardown -> set_edit_approval_requester(None)).
2026-09-03 13:04:17 -07:00
Teknium b610e603db simplify(compat): plugins/platforms+web — drop 7 re-exports + 3 aliases, repoint 1 caller + 11 tests, re-remove credential_summary
dingtalk: drop DINGTALK_TYPE_MAPPING/EXT_MAP re-exports. google_chat: card_spec_to_cards_v2 test -> .cards.
matrix: drop module-level MAX_MESSAGE_LENGTH alias (no importers). teams: drop TeamsSummaryWriter re-export
(teams_pipeline/runtime + tests -> summary_writer). wecom: drop WeComStreamExpiredError/STREAM_EXPIRED_ERRCODE/
MAX_INTERMEDIATE_FRAMES re-exports (tests -> .streaming). parallel: drop _get_parallel_client/_get_async_parallel_client
aliases (tests -> _get_sync_client). email: drop stale 'alias' comment (_esecret_int is the only name).
photon: re-remove credential_summary() (shim-only, cb9b7c36f3); its no-leak test now drives print_credential_summary.
2026-09-03 13:04:17 -07:00
Teknium ecf760db96 simplify(compat): plugins/memory — drop 15 re-exports + 1 alias, repoint 13 test callers
hindsight: drop _PORT_HEALTH_GRACE_ENV/_sanitize_bank_segment facade re-exports (tests -> .embedded/.settings).
honcho: drop 5 tool-schema re-exports, _credential_fingerprint (-> client_cache), _is_auth_error (-> session_auth),
and the _redact_tokens alias (callers renamed to redact_tokens).
openviking: drop 5 _setup re-exports; tests call openviking_module._setup.* directly.
2026-09-03 13:04:16 -07:00
Teknium 9e3df66e95 review-fix(locks): keep the lock-free cache read only for manual codex /compress (BASE exception)
BASE's _compress_codex_app_server_session was the single cache-peek site that read
_agent_cache without a lock when _agent_cache_lock was None (tests/gateway/
test_codex_hygiene_compaction.py::_slash_host relies on it). Every other BASE caller returned
None without a lock. Model that as an explicit lockless_fallback=True at that one call site.
2026-09-03 12:42:38 -07:00
Teknium 6e2796b82d review-fix(locks): restore BASE fail-loud lock-slot reads and _is_openai_client_closed truth table
Adolanium review §3 (Medium/Low). BASE shape at every steer/redirect/persist/agent-cache
lock site was: lock = getattr(obj, "_x_lock", None); if lock is not None: with lock: <direct
attribute read>; else: <getattr fallback for object.__new__ test stubs>. The refactor rewrote
several of these as 'with getattr(...) or nullcontext(): getattr(slot, None)', which (a) turned a
missing slot under the lock from a loud AttributeError into silent None and (b) in the gateway
peek helper read the cache without any lock when the lock attribute was absent.

Restored BASE semantics at:
- agent/agent_runtime_helpers.py::_requeue_pending_steer
- agent/interrupt_control.py: steer, redirect, clear_interrupt, _has_pending_redirect,
  _drain_pending_redirect, _drain_pending_steer (new _ic_slot helper: direct read under lock)
- agent/session_persistence.py::_persist_lock (explicit None check + BASE rationale)
- gateway/slash_commands.py::_cached_agent_for (BASE callers read ONLY under the lock; no lock -> None)
- gateway/run_agent_cache.py::_evict_cached_agent (BASE: self._agent_cache direct under lock)
- agent/client_lifecycle.py::_is_openai_client_closed: BASE body + docstring verbatim (outer
  is_closed first; inner _client.is_closed only when _client exists; else False)
- agent/stream_delivery.py::_ensure_stream_writer_state: restore BASE rationale that the lock is
  created unconditionally in agent_init (_STREAM_STATE) and the lazy path is stub-only

A/B (/tmp/rf/rev/ab_lock_fallbacks.py, ab_client_closed.py) is byte-identical BASE vs HEAD on
the reviewer's inputs + edge cases. tests/agent/test_lock_fallback_base_semantics.py pins it
(7 of 25 cases fail on the pre-fix tree).
2026-09-03 12:29:52 -07:00
ethernet 283c4f058c fix(packaging): derive root py-modules from the tree in setup.py
The static py-modules list in pyproject.toml drifted from the source
tree each time the root layout changed. An installed wheel then
raised ModuleNotFoundError on import: hermes_state_common, and every
gateway or CLI start failed. The list missed hermes_state_holders,
mini_swe_runner, and the 15 new hermes_state_* modules from this
branch.

setup.py now derives py_modules from the source tree at build time.
setuptools package discovery sees only directories with an
__init__.py, so root single-file modules need py_modules in every
wheel build. setup() kwargs merge with pyproject.toml, and setup.py
is the only legitimate wheel or sdist builder, so the derived list is
the single source of truth.

The nix build is the only wheel consumer. Its source filter keeps
every root .py file, so the build sandbox derives the same set as the
checkout. Editable installs do not read py_modules: build_editable
never runs bdist_wheel.

Verified: wheel and sdist built with HERMES_NIX_BUILD=1 carry all 38
root modules. The guard still blocks builds without the nix env var.
The sealed uv2nix venv from nix build .#default imports
hermes_state_holders, hermes_state_sessions, mini_swe_runner, and
hermes_startup_watchdog.

(cherry picked from commit e23d467b523d50ad088b89b578017996709589cf)

Invariant test updated to pin the derived list (no static py-modules; every root module packaged code imports ships).
2026-09-03 12:22:50 -07:00
Teknium c65aebccf0 review-fix(conversation-loop): finalize_turn binds stream-recovered final_response before fallible persist sub-steps (BASE order)
On BASE the #95514 stream-recovery rebound final_response inline, before
the tail-closing / persist-override / micro-compaction / _persist_session
calls inside the guarded persist try-block, so a raise in any of them left
the caller with the streamed text (cleanup_errors reported the failure).
HEAD's _close_transcript_tail returned the recovered value only on normal
completion; an exception in the same block dropped it and the user-visible
answer regressed to the pre-recovery '' (then rewritten by the explainer).

Split the helper into _drop_transcript_scaffolding / _recover_final_from_stream
/ _close_transcript_tail and rebind final_response in the guarded step as soon
as it is computed, restoring BASE ordering. Adds a regression test.
2026-09-03 09:54:04 -07:00
Teknium c447b9c80c review-fix(compute-host): turn-protocol tests against the real (non-seeded) frame path
ethernet8023: the seeded E2E (turn.start -> deltas -> turn.end with history_version/
message_count) was cut to hello+bogus-frame while ComputeHost._run_real_turn stays live.
New suite drives server._sessions + _run_prompt_submit through the host: turn.started,
message.delta rpc frames, turn.end identity (history_version=1, message_count=2), sid-
required/session-busy turn.error, stale queued-generation -> interrupted, live interrupt
frame ack + interrupted turn.end, and the explicit-stop compat matrix ported off HostSession.
2026-09-03 09:53:35 -07:00
Teknium 34abf954bd review-fix(public-api): restore get_session_activity, latest_user_message_row_id, resolve_multiple_toolsets, has_provider, nous_token_has_billing_scope, curated_models_for_provider, clear_edit_approval_requester + tests
All public on BASE 63279301bc, dropped by the simplify refactor (their tests were deleted or
rewritten to the replacement API). Restore each with BASE signature/body as a thin wrapper over the
surviving implementation, and restore the tests at the original call sites: test_message_reactions
again asserts the role=user contract (a newer assistant message is never the default target);
test_hermes_state / test_watchdog_review_76354 go back to get_session_activity(); toolsets, acp auth,
edit_approval, billing-scope and curated-models tests restored/extended.
2026-09-03 09:40:49 -07:00
Teknium 8da430d69c review-fix(bedrock): restore call_converse_stream + classify_bedrock_error (public on main) and the 4 deleted streaming tests; pin the live agent stream recovery path
ethernet8023: IAM streaming-denial -> converse fallback and stale-ConnectionClosedError
client eviction lost all coverage. Both public helpers restored byte-identical to BASE
(with THROTTLE/OVERLOAD/CONTEXT_OVERFLOW patterns + is_context_overflow_error), the 4
tests re-added verbatim, and TestAgentBedrockStreamRecovery covers the path the agent
actually uses (chat_completion_helpers._bedrock_converse_call / _BedrockStream._fall_back_to_converse).
2026-09-03 09:39:57 -07:00
Teknium 9c7238572e review-fix(vision): restore _validate_image_url/_image_url_shape_ok (public on main) + SSRF/localhost-block tests
ethernet8023: TestValidateImageUrl (incl. localhost block) was deleted with the sync
validator. Both names restored with BASE semantics; the async validator (the live
download path) now shares _image_url_shape_ok and gets its own localhost/malformed test.
2026-09-03 09:37:18 -07:00
Teknium cb9b7c36f3 review-fix(photon): restore credential_summary() (public on main) + test_credential_summary_no_secret_leak
ethernet8023: the no-leak test was deleted alongside the function it pinned; the
display-only credential status is live in print_credential_summary. credential_summary
restored byte-identical to BASE so the leak contract is CI-pinned again.
2026-09-03 09:36:07 -07:00
Teknium 057681739f review-fix(voice): restore listen_for_speech (main's rolling-floor VAD barge monitor) + its 6 tests
ethernet8023: both documented VAD false-trip regression tests were deleted with the
function they covered. listen_for_speech is public API on main (plugins may import it),
so the body is restored byte-identical to BASE; full_duplex_listen/_BargeDetector is
A/B-verified identical to BASE over 400 fuzzed playback/speech scenarios.
2026-09-03 09:35:38 -07:00
Teknium 3a8e3a2e88 review-fix(moa): restore _relay_moa_reference_event/_moa_reference_output_allowed + quiet-reference test
BASE (63279301bc) exported agent.agent_init._moa_reference_output_allowed and
_relay_moa_reference_event (salvaged in 3dfe712384 from #67334, plugin-importable,
zero in-tree callers) and covered them with
tests/agent/test_moa_quiet_reference_output.py. The simplification PR deleted both
the helpers and the test. Restore them byte-identical to BASE.

Note: the live build_moa_facade relay in agent/moa_loop.py is unchanged. A/B
harness (/tmp/rf/rev/moa_quiet_ab.py) shows BASE's facade relay already fired
moa.reference under platform=cli/tool_progress_mode=off — the guard only ever
lived in the orphan helper. -Q is protected on both trees by cli.py nulling
agent.tool_progress_callback (_configure_quiet_agent / BASE cli.py:22308).
2026-09-03 09:35:34 -07:00
Teknium 96c104c903 review-fix(public-api): restore from_platform_entry, SessionManager.remove_session/cleanup, is_relay_media_url, SessionTurnLeaseRegistry.__len__ + tests
BASE exposed these public names; the simplify refactor dropped them (and deleted/removed their
tests) although plugins/connectors import them. Restore each with BASE's signature and body
(download() again routes its auth decision through is_relay_media_url), and restore the covering
tests ported to the new layout, plus DB-only/task-cwd coverage for remove_session/cleanup, the
zero->4096 chunking normalization for from_platform_entry, and len() on empty/populated registries.
A/B vs 63279301bc: /tmp/rf/rev/ab_publicapi.py identical output on both trees.
2026-09-03 09:32:40 -07:00
Teknium 58dca1be51 review-fix(gateway): pin split_markdown_atoms empty-fence atoms to main
main's 'and len(current_lines) > 1' fence-close guard is dead (the opening
fence is always already in current_lines), so dropping it is byte-identical
for every input (A/B: 11 empty/nested fence shapes). Add a parity test that
asserts main's atoms for empty ``` blocks so the boundary stays pinned.
2026-09-03 09:32:38 -07:00
Teknium febbe6ae12 review-fix(computer_use): restore per-axis scroll coordinate semantics
main guarded scroll's x/y independently (coordinate=[null,100] -> y=100),
while click treats a coordinate without x as no point. The shared _xy()
applied click semantics to scroll; split out _scroll_xy() with main's
per-axis guard and pin both behaviors in a test.
2026-09-03 09:32:38 -07:00
Teknium 0c62b297c6 review-fix(mcp): preserve MCP SDK 2 input schemas in the schema-cache write-through (salvaged from #91451 / #102129)
`_write_schema_cache` in the extracted owner tools/mcp_tool_registration.py
read `t.inputSchema` with a bare camelCase getattr. mcp 2.0 renamed the
model field to `input_schema` and kept `inputSchema` only as a serialization
alias, which pydantic does not apply to attribute access, so on SDK 2.x the
cache persisted `"inputSchema": {}` for every tool and a `lazy: true` server
registered from that cache was advertised to the model with all parameters
stripped. Pre-existing on BASE (tools/mcp_tool.py); fixed with the canonical
`mcp_field(t, "input_schema", "inputSchema")` helper, as both PRs do.

Adds the real-SDK regression test from #102129 (genuine `mcp.types.Tool`
driven through `_register_server_tools` -> cache write ->
`_register_from_cache_sync`), ported to the new module layout and isolated
from the module-global registration state.

Co-authored-by: mzkarami <mehrzad.karami@gmail.com>
Co-authored-by: lijinxiao1982 <120761624+lijinxiao1982@users.noreply.github.com>
2026-09-03 09:32:31 -07:00
Teknium eaf10c7e78 review-fix(packaging): declare every hermes_state_* root module in py-modules
The hermes_state split added 15 root-level sibling modules but
[tool.setuptools] py-modules still listed only the old set, so the built
wheel/sdist (and the uv2nix sealed venv) failed at 'import hermes_state'
with ModuleNotFoundError. hermes_state_holders and hermes_state_registry
were already missing on main (registry is imported by gateway/run.py and
hermes_state.py).

Add an invariant test that parses pyproject and walks every import in the
packaged root modules and packages, requiring any import that resolves to
a root-level *.py to be declared - so the class of bug is caught by CI
rather than by a broken install.
2026-09-03 09:31:01 -07:00
Teknium aec50eb597 fix(live-qa): nous_subscription managed table indexes only selectable features (KeyError 'modal' crashed status/tools/dashboard toolsets); restore PluginManager public API (has_portable_mcp_servers, get_telegram_handler_factories, register_approval_transport), extract_json_candidate public name, canonical_whatsapp_identifier main semantics 2026-09-03 08:16:44 -07:00
Teknium 022785a541 Merge origin/main (63279301bc): reasoning-mandatory 400 recovery folded into turn_recovery/error_classifier/models_reasoning_caps 2026-09-03 05:36:21 -07:00
Teknium 63279301bc fix(reasoning): retry after a mandatory-reasoning 400 resends the user's own effort
The retry must land on the same provider cache key as every prior request
in the session. Discard only the one-shot continuation disable and send
agent.reasoning_config verbatim; a config that is itself a disable is
omitted (that session never sent anything else, so nothing warm is lost).

Live: user effort=high, ephemeral disable → 400 → retry carries
{enabled: true, effort: high}.
2026-09-03 05:20:11 -07:00
Teknium f6bd1633f7 fix(reasoning): GLM-5.3 on Nous/OpenRouter no longer 400s when thinking is disabled
Reasoning-mandatory routes answer reasoning: {enabled: false} with HTTP 400
"Reasoning is mandatory for this endpoint and cannot be disabled". Hermes
sends that disable for /reasoning none, agent.reasoning_effort: none, and the
one-shot thinking-exhaustion continuation override (which GLM-5.3-flash
triggers on its own). The Nous profile's catalog guard swallows the disable
only when its per-process capability cache already says mandatory; a gateway
that warmed the cache before the route flipped kept sending it, and the 400
was classified as a non-retryable format_error that aborted the turn.

- error_classifier: new reasoning_mandatory reason (retryable, no fallback,
  no compression), matched before the request-validation branch.
- conversation_loop: one-shot recovery — set agent._reasoning_disable_rejected,
  queue a catalog refresh for the provider, retry.
- chat_completion_helpers: _reasoning_config_for_wire drops every disable
  (configured or ephemeral) once the route has rejected one.
- hermes_cli/models: refresh_reasoning_caps_async(provider) forces a
  background re-fetch of the Nous/OpenRouter catalog.
- openrouter profile: omit a disable when the catalog marks the route
  mandatory (parity with the Nous profile).

Live: z-ai/glm-5.3-flash on the Portal with a poisoned mandatory:false cache.
Before: turn aborted with the 400. After: one retry, thinking stays on, turn
completes.
2026-09-03 05:20:11 -07:00
Teknium 5dcd3b2620 ci: ASYNC220 ignore follows Popen to run_shutdown.py; windows-only tests patch the defining module; whatsapp taskkill kwargs explicit 2026-09-03 05:12:25 -07:00
Teknium 8ae52dd93c Merge simp/r3-30 (late tail) into hermes/simplify-codebase 2026-09-03 05:04:08 -07:00
Teknium e73598a3c3 Merge branch 'simp/r3-36' into hermes/simplify-codebase 2026-09-03 04:34:20 -07:00
Teknium e010327ff7 Merge branch 'simp/r3-34' into hermes/simplify-codebase 2026-09-03 04:34:19 -07:00
Teknium 3bff2a581d Merge origin/main (562ee8ab76): fold usage-less empty-response guard, unterminated tool-call strip, skills-first memory guidance into the simplified modules 2026-09-03 04:31:46 -07:00
Teknium 562ee8ab76 fix(prompt): agent routes task-learned knowledge (incl. user preferences/corrections) to skills; memory is the every-session exception
The memory guidance led with 'Save proactively' and the memory tool schema
ranked 'user preferences & corrections' as top priority, while the skills
nudge was a conditional 'offer to save'. In practice that asymmetry made
the agent end sessions writing memory entries (fighting a 2,200-char
budget) and skip updating the skill it had just used, even though the
procedure was the reusable artifact. Both surfaces now state the same
rule with skills first: what you learn doing a task, including the user's
preferences and corrections for that kind of work, goes in the task's
skill; memory is only for facts that apply to every session.
2026-09-03 04:02:35 -07:00
Teknium f2e82f6208 refactor(tui_gateway): hosted-room X1 — inline one-site wrappers, drop dead PeerRunsHTTPError attrs, compact lease/cancel plumbing 2026-09-03 03:59:09 -07:00
Teknium 7b72fd1247 fix(agent): stream cut mid tool-call markup no longer persists as assistant prose (#101899)
GLM-style models serialize tool calls as XML in the text channel; when the
stream drops mid-serialization with finish_reason=stop, the orphan
<arg_key>/<arg_value> fragment (or a bare unclosed <tool_call> opener)
matched neither the complete-block stripper nor the partial-stream guard
and was stored and displayed as ordinary assistant content.

strip_think_blocks (storage boundary) and the CLI display copy now strip
an unterminated block-boundary tool-call opener, or any line carrying
stray argument markup, to end of text. The response then reads as empty
and flows through the existing empty-retry path. Complete blocks and
inline prose mentions are unchanged.
2026-09-03 03:42:15 -07:00
Teknium c3e9defd18 fix(agent): trim usage-less empty fix to guard + observability
Drop the loop-side '(empty)' rewrite (the turn-completion explainer already
owns that at delivery, and gateway/desktop match on the sentinel) and the
extra token-count persistence. Keeps: usage-absent empty streaks arm the
deterministic fast-fail after two attempts with no content or reasoning,
and every completed API call logs even when the provider omits usage
(#101898).
2026-09-03 03:41:52 -07:00
fangliquan 3755dca7d6 fix(agent): handle usage-less empty responses 2026-09-03 03:41:52 -07:00
Teknium 0071ba9965 Merge origin/main (561b053f79) into simp/forwardport: forward-port 220 main commits into the simplified tree 2026-09-03 03:31:03 -07:00