- tool results carry a budget reminder near limits; text before the last tool
call is dropped as report when cut by budget/abort
- usage persisted when a run suspends, so waiting sessions show spent tokens
- task page shows budget_exceeded via business.outcome instead of completed
- budget-exceeded decisions merge by root cause (decision / schedule / goal
template); new summaries added to investigate allowlist
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- three read-only summary capabilities (38 registered); task.inspect points
download task ids to download.inspect
- probe queries reject empty / single-letter / wildcard-only terms; externalId
documented as string; every description says when to use it
- download.plan requiresApproval uses coveringGrant (shared with authorize)
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- browser.screenshot returns the page as an image content block (≤1280px wide, JPEG quality
configurable) for vision-capable upstreams; non-vision upstreams get a redacted text snapshot.
Login pages / password fields are marked sensitive and withheld unless a
browser.sensitive_screenshot grant exists and allowSensitive is passed.
- browser.click_at / type_at / scroll / wait act in screenshot pixels; the server converts via
viewport/DPR, requires automation control and an allowlisted host, and writes a thumbnail
evidence file (DATA_DIR/tmp/evidence) referenced from agent_run_event.
- browser.solve_challenge runs a bounded child agent_run (vision caps only, maxSteps ≤ 12,
budget), judges success from page facts, stops immediately on 2FA / SMS / device confirmation,
and otherwise moves the session to waiting with evidence and a takeover link.
- PiAgentRuntime passes modelImage outputs as pi-ai image content when model.input has image;
the ledger never stores image data; coordinate actions use per-invocation idempotency keys.
- Tests: image codec, runtime image content, E14 e2e (login solved, 2FA stop, step limit,
sensitive withheld). Real camofox 1.17.0 check on example.com recorded in the runbook.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Auto-fix of 602 string literals across 74 files (half-width punctuation next to
Chinese, parenthesised Chinese, quotes); biome reflow; e2/e6 assertions follow
the new full-width summaries. check:copy now runs with --strict in ci.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- agent.plan emits plan_updated; system prompt asks for a plan first and updates while executing
- agent.propose_options registers a decision and suspends the run (waitForFact, no worker slot);
the answer or timeout (recommended/stop) resumes it as the tool result
- console-session runs get inline approvals for plan-bound capabilities: approve replays the call
with the approval, reject fails it with the reason and records it on the decision
- proposeFirst blocks execute capabilities until an option is chosen; E12 covers all paths
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- GET /auth/totp/setup returns the pending secret and otpauth URL (created
on first call, reused on refresh) for the console to render as a QR code.
- Enabling TOTP issues 8 one-time recovery codes (argon2id hashes in
admin_recovery_code); login accepts recoveryCode in place of totpCode,
disable accepts either, and recovery_codes regenerates with the password.
- Passkeys via @simplewebauthn/server: register/list/rename/delete under
the console session, public login options/verify that reuse the login
rate limit and session cookies. Challenges are one-time rows in
webauthn_challenge; RP ID and origins come from WEBAUTHN_RP_ID /
WEBAUTHN_ORIGINS, falling back to the request host. Signature counter
regressions are rejected. SessionInfo gains passkeyEnabled.
- --role rotate-master-key re-encrypts the secret table from
SECRET_MASTER_KEY to SECRET_MASTER_KEY_NEW one row per transaction,
skips rows already on the new key (resumable, swap keys to roll back),
logs progress and audits secret.master_key_rotate. Secrets now record
the master key fingerprint in kdf_params.kid.
- Runbook: recovery codes, passkey configuration, rotation steps and
rollback.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
agent.await and agent.spawn{await:true} now return immediately when
children are still running: the session closes, runAgent registers a
wake signal in agent_run.await_wake and waits with waitForFact without
holding an io/agent worker slot. A child that settles (terminal or
waiting) signals the parent's registered fact, so the parent resumes at
once instead of polling; the suspended call replays and the child
report becomes the tool result. agent.spawn is registered replay-safe;
a suspended spawn is recorded as failed in the ledger so the replay
re-runs it against the same child.
Migration 0037 also adds the admin_recovery_code, admin_passkey and
webauthn_challenge tables used by the account work.
E11 asserts the parent is suspended with no run held by any worker
while it waits, and covers spawn{await:true} end to end.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
/browser/providers, /browser/sessions (open/close/list), navigate/snapshot/screenshot/
click/type/scroll/select/extract/download and control acquire/release. Human input
needs an active lease; expiry or captcha moves the session to waiting with a decision;
release forces a fresh snapshot. Downloads are staged under DATA_DIR/tmp, verified by
infohash and submitted through POST /acquisition-plans. Capability count 21 -> 25.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- include/exclude sync to a per-subscription private rule group (rule_group.hard, migration 0036);
evaluation merges chain hard conditions; private groups hidden from GET /rules, soft-deleted with
the subscription and rebuilt on resubscribe
- MP resolution / quality / effect map to presets written onto the subscription's own slots
(default slot template on create); UHD / Atmos go to the private group include
- downloadSavePath mapped to the downloader path and passed as qB savepath; libraryRootId used
for the first organize when the work has no path binding; keywords appended to active search
- POST /subscriptions/{id}/reset releases stale in-flight plans, clears deferrals, re-evaluates
- GET /subscriptions/calendar from episode air dates; refreshMetadata writes TMDB season
episode air dates for tracked seasons
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- site.list / site.health_check / site.refresh_login / site.user_data capabilities
over the W9-A site center services; outputs whitelisted (no secret refs,
cookies, passwords, login names); execute kinds need automation grant.
IDs use snake_case (CapabilityId regex); registry count 17 -> 21.
- settings.organize.minSizeBytes (default 50 MiB); file-manager identify /
organize and planImport take request > setting > port default.
- tests and runbook notes (sites, files).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
When children are still running, agent.await records the wait, closes
the session (the tool call stays pending in the JSONL) and drive returns
suspended. runAgent waits with waitForFact for the children to settle,
the deadline, or cancellation, then reopens the session; agent.await is
replay-safe, so resume replays it and the model gets the real result.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
pi.provider / baseUrl / model / thinkingLevel / apiKeySecretRef are now
DEPRECATED_SETTINGS_FIELDS: kept in the schema for stored JSON, dropped
from PATCH /settings and from the settings sections, never read by the
server. runAgent fails with a clear message when no upstream is set.
The agent demo gate checks the default upstream provider; e2e seeds a
fake upstream as default. OpenAPI summaries say Agent upstream.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- agent_run gains depth / spawn_reason (0029); agent_run_event types
subagent_spawned / subagent_event / subagent_completed
- agent.spawn (propose, medium, replay never): allowlist ⊆ parent, mode not
above parent, budget carved from parent's remainder, idempotent per run+input;
settings.pi.subagents {enabled, maxConcurrent, maxDepth} (subagentsEnabled kept as alias)
- child runs claim a concurrency slot under an advisory lock (waitForFact while
queued); agent.await polls business tables without calling the model
- parent abort cascades to all descendants; parent runtime budget counts
descendant usage; child events mirrored into the parent stream by parent seq
- authorize caps every child call by all ancestors' allowlists/kinds
- session page shows parent link, inherited info and the subagent tree
- e2e E11; runbook section on subagents
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- modules/schedules: registerScheduledJob({key, defaults, run}) registry; defaults are
synced into scheduled_job without overriding user-changed enabled/schedule
- scheduler runs only on the io worker, one instance via pg_try_advisory_lock held on a
dedicated connection; due = enabled and next_run_at reached, or run requested;
jobs run in parallel, never re-entrant; interval next = finish + period, cron next match;
run-now does not move next_run_at and merges duplicate requests; skip advances one period;
history trimmed to the last 50 runs; interrupted runs marked failed on takeover
- agent_check jobs: trigger condition checked first, goal template rendered, runAgent
enqueued via startRun(trigger=schedule) with allowlist and budget, one run per time slot;
allowRecover uses the autoInvestigate grant check; results harvested back into the run
history and optionally reported as an info decision; 3 built-in checks, off by default
- sources-identity (feeds.poll, sites.snapshot, identity.review), acquisition-organize
(downloads.sync, cache.threshold, downloads.orchestrate, slots.reevaluate,
subscriptions.search, subscriptions.metadata, library.reconcile) and file-manager
(trash.purge) register jobs instead of running their own setInterval; default
intervals unchanged; backup registered as an unconfigured placeholder
- endpoints under /schedules (admin, console session only)
TODO: pi refreshPiTokens and auth-settings session cleanup still own their timers.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- POST /files/probe writes media_asset.media_info, so scope organize.read -> organize.execute
- e2e e6-w4a-endpoints -> e10-w4a-endpoints (number clashed with e6-auto-recover)
- trash: reusing a consumed approval / retired planId is rejected and the file stays
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- downloads.ts observe: re-read download_task FOR UPDATE inside the tx and merge
from the locked row; qB sync could overwrite imported back to importing when
verifyAndImport committed between the sync read and write (E8 flake).
- agent finalize: accept summaryPrefix so the recovery verdict is written with
the terminal status in one update (E6 read a budget_exceeded run before the
"未恢复:…" annotation landed); recover-summary step kept, now idempotent no-op.
- organize.ts: drop local slotComplete and the once-subscription completion
write; completion/history are decided only by applySlotEvent(verify_passed)
(W5-C follow-up, removes double write). Manual slot organize uses the shared
subscriptions.slotComplete.
- vitest: server split into unit (parallel) and pg (file-serial, groupOrder 1)
projects; packages/tasks file-serial; explicit worker concurrency in e2e.
- docs: W6 stabilization record in 04-v0.1-acceptance.md.
Test expectations unchanged. Full `deno task ci` x3 green: 369s / 341s / 337s.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- completion by season episode_total ledger (applySlotEvent verify_passed hook overrides organize's
complete): unknown total keeps tracking and records suspected-missing; pending never completes;
once / continuous current-season completion writes subscription_history
- evaluation: missing = release units ∩ episodeRange − covered − in-flight; series slots may hold
several unit-disjoint in-flight plans; overlapping/busy releases are deferred (slot_deferral) and
re-evaluated when a plan ends (next_evaluate_at due) or on the periodic slot re-evaluation
- per-unit upgrade via slot_unit ledger; site scope (params.siteIds); site facts in evaluation context
- season packs: after submit read qB files and setFilePriority(0) for covered/in-flight episodes;
whole pack with recorded reason when undeterminable; removed downloads release their reservation
- searchSubscription task (title/aliases × enabled sites via liveSearch, snapshot + rate limiter):
on create, every searchIntervalHours (default 24h) and POST /subscriptions/{id}/search
- refreshMetadata writes season totals (only grows, floored by covered, lock respected) and turns
metadata_pending subscriptions active; plan loop refreshes tracked series every 6h
- endpoints: subscription progress, episode-total, subscription-history; e9 continuous-series e2e
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- POST /organization-records: sourcePaths must resolve (realpath) under a
registered library_root or downloader path-mapping host path; `..` and
symlink escapes rejected (400), dryRun included. Move requires an admin
session, or an API key with an automation_grant covering the entity or a
matching one-time approval (consumed on use), else 403. Manual Move from a
managed download_task with unsatisfied H&R returns 409 and opens a
decision; never silently downgraded to Copy.
- Series slot completion: total episode count is unknown while units are
provisional, so the slot is not satisfied and the subscription keeps
tracking; coverage is registered per actual file (creating provisional
episode units), and already_in_library files count as verified presence.
- POST /engine/runs/{id}/retry (admin): re-enqueue by original task name and
raw input with idempotency key retry:<id>:<n>, audited; retry button on
the run detail page.
- E7 regular series e2e, E8 AT24 tag removal / task disappearance e2e.
- All workspace members and root set to version 0.1.0; CHANGELOG.md.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- settings.agent.autoInvestigateMode investigate|recover, investigateMaxAttempts, recoverVerifySec
- new capability download.recheck (execute, replay never): qB force recheck / resume, no source swap/delete
- autoInvestigate moves to agent/investigate.ts: recover mode runs execute with a recovery allowlist
only when an active grant (published policy if bound) covers the recovery actions; otherwise degrades
to read-only investigation and leaves suggestion + recheck option on the decision
- runAgent verifies recovery against business facts (download_task/task/source_release) before finalize;
verified -> decision resolved by system:auto-recover, run summary '已恢复(系统核实)'; otherwise stays
awaiting_user; per-root-cause attempt cap
- e2e E6: missingFiles -> download_failed -> recheck -> verified; no grant degrades; swap rejected;
dedupe; attempts exhausted; budget exceeded keeps attention
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The business row settles before the engine marks the workflow completed; under CI load the
window is a few hundred ms and the direct assertion flaked.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- Capabilities: all 14 registered (AO: inventory.query, download.plan/submit, organize.plan/execute,
policy.preview/apply; agent: task.inspect/retry/cancel); startup asserts count == 14.
Registry injects server-side authorize (automation_grant by action/scope/policy version/budget,
one-shot approval bound to plan id + payload hash + revision + expiry; model claims ignored;
automation switch), audit_event, server idempotency keys and a capability_invocation op ledger
(migration 0021) used for crash reconciliation. download.plan creates draft plans only;
download.submit promotes draft -> reserved after authorization.
- Agent session API: POST/GET /agent/runs, GET /agent/runs/{id} (events by seq, messages, report),
POST .../messages (wake), POST .../abort; automation-grants, approvals, capabilities invoke, tasks.
- runAgent task: Pi or fake provider from settings (apiKey via secrets), events persisted to
agent_run_event then published, resume + ledger reconciliation when pendingOperations > 0,
waits for goal-linked downloads/imports and reports from business tables (submitted != imported).
- Background triggers: decision.created -> bounded plan_only investigation (dedupe per decision);
decision.updated(resolved) -> resolveDecision executes chosen option.
- W2 closure: all OpenAPI routes registered in contracts api.ts (+ subset test); TMDB key from
settings secret_ref with env fallback; PUT /sites/{id}/credentials documented as alias of
/site-accounts; completion events documented as existing status events.
- e2e (real PG + OpenWorkflow + testkit mocks): RSS->library, agent delegation with grant/deny,
background continuation, worker kill -9 recovery (single qB add), auto investigation dedupe.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
