* fix(agents): do not label finish_reason error as LLM timeout
Bare provider finish/stop reason `error` is a completed failure, not a
hang. Classify it as server_error so failover stays eligible, and keep
the provider signal in user-facing copy instead of rewriting to
"LLM request timed out." Abort/network/malformed finish reasons remain
timeout-like.
Fixes#109218.
Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>
* test(agents): assert concrete finish_reason error copy
Avoid optional-string chaining that fails check-test-types (TS18048).
Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>
* fix(docs-sync): declare pruneOrphanLocaleDocs in .d.mts
Root test types import the export from docs-sync-publish.mjs, but the
declaration file omitted it after the recent docs-sync change on main.
Unblocks check-test-types for this PR and current main.
Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>
* fix(scripts): register finish-reason proof script for knip
* docs(agents): classify bare finish_reason error as server_error
Align model-failover docs with the #109218 classifier: provider-completed
bare error stop/finish reasons are server_error (500), not timeout.
Expand the controlled-provider proof to show stream map, failover status,
user copy, and model-fallback eligibility.
Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>
* fix(scripts): drop unrelated docs-sync .d.mts drift
check-guards fails: pruneOrphanLocaleDocs is declared in
docs-sync-publish.d.mts but not exported from the .mjs. That
declaration was not part of the finish_reason error fix; restore
the main contract.
Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>
* test(agents): cover finish reason error transport
---------
Signed-off-by: Sebastien Tardif <sebtardif@ncf.ca>
Co-authored-by: Altay <altay@hey.com>
* feat(gateway): session observer digests over the utility model
* fix(gateway): split session-observer modules and satisfy ci gates
* fix(gateway): observer reads session entries without materializing agent state
* fix(agents): drop silent reasoning-only replay turns
Prevent NO_REPLY cleanup from leaving signed reasoning that merges into the next assistant tool-use turn and bricks strict-provider sessions.
Related: #99620
Co-authored-by: SunnyShu <shu.zongyu@xydigit.com>
* docs: refresh transcript hygiene map
Regenerate the docs map after documenting silent reasoning-only replay cleanup.
Co-authored-by: SunnyShu <shu.zongyu@xydigit.com>
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
A Gateway timeout or closed connection now fails the command with an
actionable stderr hint instead of silently re-running the whole turn
embedded under a fresh gateway-fallback-* session. The silent fallback
could double-execute side effects (the Gateway may still finish an
accepted turn), returned context-free answers to --session-key callers,
and ran with the CLI host's local config. --local remains the only
embedded execution path.
Also deletes the resultMetaOverrides plumbing (the fallback was its only
writer) and the fallback marker fields added in #111645.
* refactor(ai): invert plugin coupling behind the transport host port
* fix(ai): queue custom transport registrations until the host is configured
* refactor(ai): remove relocated transport sources from src/agents
* fix(ai): source core stream types from canonical packages and fix tarball fixtures
* fix(ai): invert plugin transport host wiring
* fix(ai): harden managed transport projection
* test(ai): register synchronous stream in transport mock
* fix(ai): lazily install transport runtime host
* fix(ai): preserve completion compat detection
Channel plugins declare package env triggers in manifests, and ambient
environment variables count as "channel configured". A dev gateway
(gateway run --dev) inherits the operator's shell, so real channel
credentials silently configured channels and could connect development
instances to live services (observed with reef and a Telegram-range
connection during isolated stress testing).
Dev mode now drops presence signals whose only sources are env or
manifest-env across activation planning, auto-enable, autostart, health
recovery, readiness, reloads, and startup warnings. Explicit
channels.<id> config still works, --dev-ambient-channels restores the
old behavior, and startup logs the suppressed channel ids once. Non-dev
gateways are unchanged.
Support ClickClack v1 setup claim URLs for split-origin and path-mounted deployments while preserving legacy setup flows and private transport overrides.
Refs #111919.
Prepared head SHA: c13ce82e38
Co-authored-by: Shakker <165377636+shakkernerd@users.noreply.github.com>
Reviewed-by: @shakkernerd
Final sweep of the media-placeholder program: with all eleven producers
migrated, the consumer regexes are dead input and are hard-deleted —
media-understanding token un-mixing (bodies are pure user text now),
the persisted-user-turn placeholder drop, the session-memory strip, and
the inbound-event mediaPlaceholder field with its replacement branch
(an intentional removal of shipped SDK surface per the program's
hard-delete decision: plugins still minting placeholder bodies render
them as literal text until updated). formatMediaPlaceholderText remains
the single owner of the <media:kind> vocabulary, used only for
text-only carriers. Legacy tokens in pre-migration history remain
ordinary text.
Route automatic dashboard titles through the effective session model and auth profile while preserving explicit cross-provider utility ownership and the existing generic title fallback.
* fix(clickclack): split API base and refresh discussion provider
* chore: regenerate bundled channel config metadata and docs map for the apiBaseUrl key
* feat(anthropic): import Claude Desktop custom groups
* fix(anthropic): resolve Claude Desktop groups from LevelDB entries
Group discovery regexed whole decompressed blocks, so matches were not
attributable to a Local Storage value and byte order decided the winner.
On a real 5.4MB store that mislabelled 59 of 159 sessions, surfacing a
mojibake label spliced out of Snappy copy-record bytes.
Parse SSTable entries properly instead: prefix-delta keys bounded by the
restart array, newest internal sequence per user key (honoring deletions),
and record scanning confined to a single value. Values are normalized so
Chromium's UTF-16 form scans like Latin-1, and unflushed WAL writes keep
precedence over SSTables.
Prod LOC grows ~77; it buys structural correctness in place of ordering
luck, and folds the old index-only walk into one shared entry decoder.
Verified against the live store: 159 assignments, 0 mislabelled versus an
independent entry-level ground truth, 0 control-character labels.
* fix(ui): sort custom session groups ahead of project groups
Custom groups were pushed into the section list as encountered, so their
position depended on roster order rather than the documented behavior;
the existing test only passed because its fixture happened to be ordered
that way. Collect custom and project groups separately and concatenate,
and assert the guarantee with a reversed-input case.
* fix(anthropic): widen LevelDB fixture key type for test typecheck
The prefix-delta helper assigned a Buffer into a Buffer-typed accumulator
whose generic argument differed, which tsgo rejects in the test lane.
* chore(anthropic): drop release-owned changelog edit from the PR
CHANGELOG.md is generated at release time, so a normal PR must not carry
an entry. The release-note context lives in the feature commit message and
the PR body instead.
Sebastien Tardif