* feat(gateway): add sessions.files.set workspace file write with hash CAS (#104698)
* feat(ui): CodeMirror file editor panel with CAS save + clickable tool-card paths (#104698)
* fix(gateway): issue the file CAS hash only for strict-UTF-8 text (#104698)
* docs(web): document the editable file detail panel (#104698)
* fix(gateway,ui): round-trip BOM and CRLF bytes through the file editor (#104698)
* fix(ui): keep mixed-line-ending files read-only in the file editor (#104698)
* fix(ui): re-gate editability when a conflict reload returns non-editable content (#104698)
* fix(gateway): reject NUL bytes in sessions.files.set replacement content (#104698)
* fix(gateway,ui): regenerate protocol bindings, fix prod types and lint for file editor (#104698)
* feat(gateway): persist operator approvals
* fix(gateway): preserve approval ids exactly
* fix(gateway): enforce approval reviewer bindings
* fix(gateway): reconcile durable approvals with exec revocation hardening
Map #103515 semantics onto the durable lifecycle: resolutionSource and
one-shot consumeAskFallback stay process-local record facts, trusted
auto-review resolves through the durable CAS as a runtime resolver, and
lost races settle with the winner's operator source. Also: derive the
audience walker cap from the store cap, report APPROVAL_ALREADY_RESOLVED
for a resolve that loses the CAS race, document approval.get/resolve in
the protocol docs, and regenerate Swift models on the new base.
* fix(approvals): validate resolver kind
* docs(approvals): explain malformed verdict denial
* fix(protocol): regenerate approval models after rebase
* chore: leave changelog entry to release generation
* fix(gateway): preserve approval registration boundaries
* test(gateway): prove multi-device approval races
* test(gateway): keep approval order assertion stable
* docs: index operator approval architecture
* fix(protocol): bound approval declaration exports
* feat(ui): detect background tasks eagerly and show live tool activity in the chat rail
Track tool-start counts and the last tool name per task run in the task
registry, persist them on task_runs, and expose them as additive optional
TaskSummary fields. The chat background-tasks rail now loads its snapshot
eagerly so the collapsed toggle badge detects running work immediately,
and rows show a live elapsed timer, tool-use count, and the tool
currently in use (run duration for finished rows).
Issue: #104775
* fix(ui): translate background-task activity strings and refresh generated protocol models
The singular tool-use label is literal ("1 tool use") because several
locales legitimately drop the count placeholder in singular forms, which
fail-closes the placeholder guard in the i18n sync. Locale bundles are
the authenticated ui:i18n:sync output (fallbacks=0); GatewayModels.swift
is protocol:gen:swift output for the new TaskSummary fields.
* chore(ui): refresh i18n raw-copy baseline after rebase
* fix(ai/gemini): coerce numeric/boolean enum values to strings
Gemini Function Calling only accepts string enum values (see
https://ai.google.dev/gemini-api/docs/function-calling). Tool schemas
containing enum: [1,2,3] or enum: [true,false] were passed through
unchanged, causing Gemini to reject the entire request with 400
TYPE_STRING errors at 'function_declarations[X].parameters...enum[N]'.
Extend cleanSchemaForGeminiWithDefs to coerce enum and const values
via a shared helper (number/boolean/bigint -> string, drop null and
non-coercible entries, de-duplicate), and retype the containing schema
as 'string' since Gemini's validator only accepts enum on string-typed
schemas.
Adds seven vitest cases covering numeric/boolean enum, const, nested
schemas, and dedup/empty-after-coercion edge cases.
Fixes#104566
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(ai/gemini): make enum retyping property-order independent
Address @clawsweeper review on #104567:
The initial patch set cleaned.type = 'string' inline while visiting the
'enum' or 'const' key. For schemas where 'type' appears AFTER 'enum' /
'const' in the input object — a valid ordering since JSON Schema property
order is not semantic — a later 'type' pass would overwrite the coercion
via the tail 'cleaned[key] = value' branch, and Gemini would still receive
an invalid declaration.
Move the type override out of the loop: track whether any enum/const was
coerced via a boolean flag, and force cleaned.type = 'string' after the
loop finishes. This makes the fix independent of input property order.
Also add curly braces around single-line ifs to satisfy check-lint
(eslint(curly)) which failed on the previous commit.
New reversed-order regression tests added.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(ai): preserve Gemini enum schema types
---------
Co-authored-by: leonme <leonme@users.noreply.github.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* refactor(sessions): migrate runtime storage to sqlite
* test(sessions): fix sqlite CI regressions
* test(sessions): align remaining sqlite fixtures
* fix(codex): require sqlite trajectory recorder
* test(sessions): align orphan recovery sqlite fixture
* test(sessions): align sqlite rebase fixtures
* fix(sessions): finish current-main integration of the sqlite flip
Resolve the whole-store SDK removal across its owner boundary: drop the
loadSessionStore re-export and the registry whole-store wrappers, wire
hasTrackedActiveSessionRun into gateway chat, complete the
preserveLockedHarnessIds cleanup contract, flip the codex thread-history
import to storePath targets, and port remaining main-side tests from
file-store helpers to session accessor reads.
* chore: drop committed pebbles log, revert plugin-inspector bump, refresh generated docs
Remove the 1.8k-line .pebbles/events.jsonl work log from the branch, restore
the plugin-inspector advisory lane to main's pinned 0.3.10 so the supply-chain
bump gets its own review, and regenerate docs_map, the plugin SDK API baseline,
and the export-surface ratchet for the merged tree.
* feat(sessions): keep archived transcripts by default with zstd cold storage
Codex-style retention: deleting or resetting a session archives its
transcript as a zstd-compressed JSONL artifact (plain when the runtime
lacks node:zlib zstd) and keeps it until the disk budget evicts oldest
first. resetArchiveRetention now governs both deleted and reset archives
and defaults to keep; maxDiskBytes defaults to 2gb so retention stays
bounded, with archives evicted before live sessions. The cron reaper
follows the same knob instead of deleting archives on its own timer.
* fix(state): converge agent DB migration lineages and bound database growth
Merge coherence: run both structure-gated legacy memory-schema repairs
(flip-lineage drop, main-lineage identity rebuild) before the flip
migration so pre-flip v1/v2 and pre-merge flip v1/v4 databases all
converge, and hoist foreign_keys=OFF outside the schema transaction
where the pragma was silently ignored and the v1 sessions rebuild
cascade-deleted session_entries.
Growth guards: fresh agent DBs enable auto_vacuum=INCREMENTAL, WAL
maintenance releases freed pages in bounded passes (never a blocking
full VACUUM), and doctor reports state/agent DB bloat from freelist
stats.
* fix(codex): resolve the store path for thread-history import via the SDK
The supervision catalog passed the legacy sessionFile locator to the
storePath-targeted transcript mirror; resolve the agent store path with
the session-store SDK helper instead of a runtime-object seam so test
fakes and headless callers need no extra surface. Drop the obsolete
missing-session-id preprocessing case: sessions rows are NOT NULL on
session_id and upsert repairs id-less patches at write time.
* fix(sessions): fail safe on malformed disk-budget config and doctor stat errors
A malformed explicit maxDiskBytes disables the budget instead of
falling back to the destructive 2gb default the user never chose, and
the doctor bloat check skips databases whose paths stat-fail instead of
aborting doctor.
* fix(sessions): complete sqlite conflict translations
* test(sqlite): align hardening checks with maintenance
* test(sessions): inspect compressed transcript archives
* fix(tests): await session seeds and drop unused helpers flagged by CI lint
The five unawaited writeSessionStoreSeed calls raced their SQLite seeds
against the assertions, failing compact shards; the bloat probe drops a
useless initializer and the merged tests drop now-unused helpers.
* test(sessions): type legacy proof events directly
* test(sessions): align hardening contracts
* perf(sessions): read usage transcript sizes from SQL aggregates
Usage/cost scans walked every session and materialized every transcript
event just to re-stringify it for a byte estimate — the #86718 stall
class reborn on the DB. readTranscriptStatsSync sums stored JSON bytes
in SQLite without loading a single row.
* fix(sessions): re-root foreign-root transcript paths onto the current sessions dir
Restored backups, moved OPENCLAW_STATE_DIR, and rehearsal copies carry
absolute sessionFile paths from the old root; the containment fallback
kept those foreign paths, so migration read (and would archive) files in
the original root and reported local copies missing. Re-root the
canonical agents/<id>/sessions suffix onto the current dir when the file
exists there; genuine cross-root layouts still fall through unchanged.
* test(agents): seed harness admission through sqlite
* fix(sqlite): close agent db on pragma setup failure
* fix(doctor): compact and retrofit incremental auto-vacuum after session import
The migration is the sanctioned offline window: post-import compact
reclaims import churn and applies auto_vacuum=INCREMENTAL to databases
created before the fresh-DB pragma existed, so runtime maintenance can
release pages in bounded passes on every install.
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* feat: add metadata-only message audit events
* chore(protocol): restore generated swift models and config baseline after rebase
* fix(state): gate agent-db ownership check before schema version
* fix(cli): sync audit command description into the root-help catalog
* fix(audit): require destination proof before classifying outbound messages as direct
* refactor(audit): drop dead migration guard and add contract comments
* docs(gateway): add dedicated audit history page and cross-links
* test(e2e): enable direct-mode message audit in the telegram proof SUT config
* test(channels): expect declared conversationKind in durable delivery session context
* fix(audit): record the routing channel id for inbound rows from plugin channels
* fix(audit): match channel-prefixed delivery targets in the destination route gate
* fix(audit): validate explicit target kind against the destination route kind
* fix(audit): use the canonical target-prefix grammar in the destination route gate and fail closed on foreign migration tables
* fix(audit): normalize nested provider and kind target prefixes in the destination gate
* fix(audit): strip registered provider aliases and the direct kind prefix in the destination gate
* chore(docs): refresh config baseline after rebase
* feat(tooling): extend strict-ratchet lane to all remaining leaf packages
* fix(packages): burn down indexed-access debt in ai and agent-core
All 196 noUncheckedIndexedAccess errors fixed behavior-identically
(iteration, .at()/slices, DataView byte math, guarded queue peeks) plus
cleanup finds: a sparse-tool diagnostic bug, deduplicated Responses
call/item ID parsing with an honest string|undefined splitter, and a
clearer local failure for redacted-thinking blocks missing signatures.
Removes all 17 remaining non-null assertions across ai, gateway-client,
and llm-core so the assertion ban aligns 1:1 with the ratchet lane.
speech-core joins memory-host-sdk as structurally deferred (imports
src/** via plugin-sdk path mappings).
* feat(tooling): add noUncheckedIndexedAccess strict-ratchet lane
* fix(packages): make indexed access explicit across ratchet packages
Burns down all 65 noUncheckedIndexedAccess errors in the seven ratchet
packages with behavior-identical restructuring (iteration, charAt/slice,
regex-group guards, validated hextet tuple). net-policy invariant
violations now throw instead of failing open. Adds
@openclaw/normalization-core/expect (expectDefined/first/last) with
subpath export and tests. memory-host-sdk fixes kept but the package
stays out of the lane: it re-exports core src/** so the flag would apply
transitively to all of core.
* fix(lint): keep .oxlintrc.json strict-JSON parseable for extension lint wrappers
* feat: node-hosted plugins — dynamic tools, MCP servers, and skills
Nodes become declarative plugin hosts:
- node.pluginTools.update: node hosts publish plugin-registered agent tool
descriptors; gateway materializes them as agent tools executing via
node.invoke under the node command allowlist, with tools.effective
invalidation and node online/offline removal.
- Trusted paired-node descriptors: no gateway-side plugin registration
required; gateway.nodes.pluginTools.enabled off-switch (default on);
description/count caps; deterministic node-prefixed collision names.
- Declarative node-hosted MCP: nodeHost.mcp.servers (McpServerConfig shape)
starts MCP clients on the node host, publishes tools as pluginId node-mcp,
executes via built-in mcp.tools.call.v1 with per-layer timeouts, failure
isolation, and orphan-safe shutdown. No re-pairing when servers change.
- Node-hosted skills: node.skills.update publishes ~/.openclaw/skills
content (64 skills/64KB/512KB caps both sides); gateway merges them into
the skills snapshot while connected and exec host=node is available, with
node:// locators, node-prefixed collisions, disabled command dispatch,
and gateway.nodes.skills.enabled + nodeHost.skills.enabled switches.
- Security: node-supplied pluginIds cannot satisfy pluginId-scoped tool
allowlists unless gateway-registered; reserved node-mcp id requires the
core MCP descriptor shape; protocol registry kept out of public
plugin-sdk dts.
- E2E: pond harness proves publication, MCP round-trip, skills locator, and
disconnect/reconnect for all three surfaces.
* style: format node-plugin-tools test
* fix(skills): keep status loader unfiltered when eligibility is passed
skills.status started passing eligibility for the node-skill merge, which
flipped loadWorkspaceSkillEntries into filtered mode and dropped disabled
skills from status reports (QA plugin-lifecycle-hot-reload timeout). Status
now merges node skills explicitly around an unfiltered load. Also: regen
docs_map for new node docs sections; add the intentional node-host MCP
onclose suppression to the lint-suppression allowlist.
Adds a git-backed session diff panel to the Control UI, complementing the existing PR status chips.
New sessions.diff gateway method (operator.read): resolves the session checkout and returns structured per-file diffs (status, renames, +/- counts, capped unified patch) of branch + uncommitted + untracked work against the default-branch merge base. Hardened against git textconv execution and hardlinked out-of-tree content leaks; handles repos before their first commit via the empty-tree base.
Control UI renders a "Changes" panel in the chat detail sidebar with collapsible per-file diffs, hunk-gap markers, stat chips, and untracked/binary badges, gated on gateway method advertisement. Schemas additive (Swift models regenerated), 20 locales translated.
Closes#104182
* feat(webui): reintroduce opt-in AI purpose titles for tool calls
Restores the chat.toolTitles path removed in #103821, gated behind the new
gateway.controlUi.toolTitles opt-in (default false) so tool rendering stays
fully deterministic with no background model calls unless an operator enables
it. Disabled gateways answer { titles: {}, disabled: true } without loading
the completion runtime, and clients stop asking for the session.
When enabled, titles use canonical utility-model routing: an explicit
utilityModel (operator-chosen provider, like every utility task), else the
session provider's declared small-model default, honoring per-session model
overrides and auth profiles; utilityModel "" disables titles and malformed
refs fail closed — never the primary model. Tool inputs are redacted with the
tools-mode redactor before cache keys or prompts, caller ids are bounded and
never reach the model, and results cache in the per-agent SQLite
cache_entries so repeat views never re-bill.
Also completes two crestodian model-input mock factories that leaked into
sibling tests under shared-registry CI shards.
Fixes#103987
* fix(webui): redact tool-title inputs before truncation
Vincent Koc