Agent definitions gain tool_approvals: known tool -> ask/auto/deny.
Unknown tool/policy is a 400, and required+deny is a 400
contradiction. The declaration is persisted (tasks.agent_tool_
approvals, PG + migrate), echoed in GET and the creation response
(via to_summary), and resolved for registry definitions -- fixing
a latent drop where resolve_registry_agent re-parsed without new
fields. At dispatch the worker audits the declaration and fails
fast on a defensive required+deny re-check (v3 tools-gate shape,
runner never called); goal/coordinator tasks reject the field at
creation like instructions. effective_tool_approvals maps a
declaration + host global approval to the actual host answer,
with ask_honoured always False: host answers elicitation with
one global switch and carries no tool identity, and the approvals
queue is record-only with no runtime ask channel, so per-tool
interception and credential scope are explicitly NOT claimed.
Policy carries an operator-declared model_capabilities directory
and per-tier requires (tool_calling/reasoning/vision booleans,
min_context_tokens). decide_routing checks the policy tier model
in active mode: capable applies, incapable falls back to a capable
worker default, and if both fail or are unknown it returns
capability_blocked, which the worker turns into a failed task
(v3 tools-gate shape, runner never called). Shadow mode only
annotates the reason, never blocks. Explicit task and worker CLI
models bypass the check by design. Invalid requires drops only
that tier (fail-closed, like invalid tier mode). Dry-run reports
per-tier capability status and blocks on active tiers that fail
or cannot be verified.
Foundry 'platform restarts with original input' gap, after
reading the existing mechanisms: creation idempotency
(UNIQUE project_id+idempotency_key), goal checkpoints
(task_checkpoints per round, same-thread breakpoint resume
with no-progress fuse), merge/cost/notification CAS, and
coordinator child reuse already exist. The uncovered segment
is non-goal platform-interrupted tasks: auto_resume requeues
them fresh in the same worktree after clearing result/error/
interrupted_reason, so the new run gets only the original
prompt and cannot tell what the previous attempt had done.
This track adds, in worker/server layer only (no host.py or
codex thread change): task_attempts ledger, recorded by
auto_resume_interrupted in the same transaction before
clearing (attempt_no = old auto_resume_count + 1, kind
goal_breakpoint|platform_fresh, interrupted_reason,
checkpoint_round, result/error truncated to 2000 chars,
INSERT OR IGNORE keeps the first record on duplicate);
resume preamble injected in-memory only at claim for
platform_fresh reruns (previous attempt facts, truncated to
1000 chars, plus check-worktree-first and verify-before-
recreating-external-artifacts instructions; skipped for
coordinator/_resume/breakpoint tasks and when no record
exists, never fabricated); attempts surfaced read-only in
GET /tasks/{id}/timeline under existing project-read
permission. Fresh PG schema, migrations table list, and
server/worker ensure paths share one DDL.
Deliberately NOT claimed: the preamble is an instruction,
not a hard guarantee like creation idempotency or merge
CAS; the ledger does not track worktree files or external
artifacts (no worktree snapshot, executor-dependent paths;
hard dedup for those still rests on their existing gates);
manual retry paths do not go through auto_resume, so they
get no ledger record or preamble. Tests:
test_attempts.py (7 tests).
Coordinator parts gain three enforced bounds, validated in
validate_parts and carried through the coordinator_parts JSON
(no schema change): timeout_seconds (1..86400, per-attempt
wall clock from coordinator start; the coordinator poll loop
cancels only that child with the same cancel mechanism as
the global timeout, other parts keep waiting, audit
coordinator_child_bound_exceeded), on_failure fail|tolerate
(tolerated failures do not fail the coordinator, are named
in the report head, and their failed children are not
requeued on coordinator re-run while interrupted children
still are), and result_chars (0..2000, default 500, enforced
in build_report; the 4000-char report writeback cap still
applies). evaluate_outcome is the single outcome rule used
by both build_report and the worker error text, so tolerated
failures never leak into the coordinator failure reason.
Unknown part names fail closed.
Deliberately NOT added, after reading the execution layer:
per-part token/cost budget (cost is recorded only on
completion, so a running child's spend is unobservable and
creator quota is monthly, not per part), per-part round cap
(normal tasks have no per-task round cap in the worker),
and write-scope isolation (sandbox is per-project worktree,
a scope string in a prompt would be an instruction, not a
guarantee). Tests: test_delegation_bounds.py (8 tests).
New pure module cloud/routing_shadow.py: parse worker_routing
audit details (old format without suggested model parses with
suggested unknown, never guessed), evaluate per-tier routed
samples with one-task-one-sample dedup, explicit tasks counted
separately, status under the chosen model, token/USD with
unpriced tokens kept apart, and a min-20 gate per tier. The
report explicitly marks the counterfactual not_measured:
shadow suggested models never ran, so no suggested-vs-chosen
cost/quality comparison is possible. Worker audit detail now
includes suggested/chosen for future rows. GET /routing-shadow
serves the report (project read permission, truncated flag
over 10000 audit rows, cost table ensured so a fresh DB does
not 500).
Tests: test_routing_shadow.py (5 tests).
Agent definitions gain tools (websearch/imagegen/playwright,
unknown tool is a 400), stored as agent_tools JSON and returned
by GET /tasks. This is deliberately required-tools, not an
allowlist: create_thread has no tool allowlist and tenant homes
are per-project shared, so blocking other tools cannot be
claimed. Instead the worker fails fast before any model call
when a declared tool has no [mcp_servers.<tool>] section in
the tenant codex-home config.toml (checked after _tenant_dirs
seeding, so a section seeded earlier by another worker counts
as present). The gate runs only for tenant-isolated Docker
workers; local/cloud workers skip it, documented, not verified.
Tests: test_agent_tools.py (8 tests).
Registry: POST /tasks accepts agent_name, loading <name>.json
from repo-root agents/ (traversal-guarded, skill-pack re-validated);
agent and agent_name together is a 400. Persistence: tasks gains
agent_name/agent_version/agent_instructions (migrate + PG schema),
returned by GET /tasks; stored prompt stays clean. Channel: local
and docker dispatch pass --developer-instructions to host.py
dispatch -> create_thread(developerInstructions=...); run/fork do
not re-send. Cloud executor has no such channel, so the worker
falls back to the v1 marked header in the in-memory dispatch
prompt only (DB prompt stays clean). Goal/coordinator tasks
still reject instructions.
Tests: test_agent_def_v2.py (7 tests) + updated v1 suite.
Pre-dispatch routing only fills in tasks with no explicit model,
classified by structural signals only (plan > goal > default).
Explicit task model and worker CLI --model win over policy;
policy via CLOUD_ROUTING_POLICY_JSON/FILE + CLOUD_ROUTING_MODE.
Shadow mode records a suggestion without changing dispatch.
Only policy/shadow decisions emit a model_routing event, so the
existing events contract (fallback tests) is unchanged.
Fallback rules, effort/approval/sandbox/quota and provider are
untouched. Tests: test_models_routing.py (20 tests).
Long-running website monitoring: new cloud/resident.py layer on top of
the automations claim/quota model (no task state-machine change).
States running/paused/ended; pause preserves the content-hash cursor
and stops polling, resume re-checks from the cursor, end is terminal
(resume after end is 409, a new watch is required). Polling is a cheap
fetch + hash comparison; only a real change creates a queued model
task. Quota exhaustion pauses (quota_exceeded) instead of ending and
keeps the cursor. Driver is the worker run_loop round boundary;
server exposes create/list/get/update/pause/resume/end with audit.
Console UI intentionally not wired this round (see verification
summary for deviations).
- click/click_ref: pre-click allowlist gate on resolvable anchor href
(same check as navigate; miss refuses before the click executes);
click/click_ref/click_at: post-click landing URL check.
click_at has no pre-click gate by design (coordinates name no
element); JS-driven navigation, evaluate, meta refresh and popups
are documented as unintercepable with the Tinyproxy 403 backstop.
- Tenant script upgrade is now safe: only proven-unmodified scripts
(manifest match, 2a569ac known-shipped SHA-256, missing, or
current) auto-merge; tenant-modified scripts are never overwritten,
are reported as skipped, and get one deduped audit record. The
manifest never learns a tenant hash (else the next upgrade would
mistake a modified file for untouched). Idempotent reruns.
- Activation candidate = deployed Tinyproxy programming set encoded
with leading-dot entries (plain entries would regress
docs.python.org); npmjs.org per the deployed filter, not npmjs.com.
Host activation itself is NOT performed this round: only the plan,
validation tests and rollback are delivered.
Tests: 28 new (click 13, upgrade 8, activation 7), all fake/no-network;
98 browser-related tests pass including prior suites.
(cherry picked from commit e4a93f5748ec815f163d0bc096ff4dc4b8c651f5)
- Multi-tab: browser_new_page / browser_list_pages /
browser_switch_page / browser_close_page in playwright_mcp.py;
each new page gets a fresh Playwright context, so cookies and
storage are isolated between pages. Single-page tools act on
the current page; switching to a missing index is an error,
not a crash; closing the last page is an error (use
browser_close).
- Navigation allowlist: BROWSER_ALLOWED_HOSTS (comma list,
leading-dot suffix match, case/port-insensitive) checked in
_check_allowed_url before launching Chromium; unset (or empty)
means no check, the old behaviour. file:// always allowed;
localhost/127.0.0.1 have no special treatment.
- Close cleanup: browser_close closes all pages and best-effort
removes cwd browser-live.png / browser-live.json; custom
screenshot files are untouched. list_pages needs no launch.
- Egress filter: 10 programming-source Tinyproxy BASIC rules
(python.org, nodejs.org, go.dev, pkg.go.dev, rust-lang.org,
developer.mozilla.org, crates.io, stackoverflow.com,
cve.mitre.org, nvd.nist.gov); already-present github.com /
pypi.org / npmjs.org not duplicated; searx.diante.me stays
exact-only, no diante.me wildcard.
- Allowlist worker wiring: _docker_task_context now passes a
set, non-empty BROWSER_ALLOWED_HOSTS from the worker env into
[mcp_servers.playwright.env] in the tenant config, so the MCP
process actually sees it (container -e alone does not reach
MCP subprocesses, 2026-10-30). Unset/empty writes nothing and
never overwrites; the value never goes into logs; it is not
added to the websearch env block or the container env.
ensure_playwright_mcp_env now returns False without rewriting
when the result would be identical (canonical join: one blank
line before the block), so a second prep leaves an unchanged
tenant config alone; other sections untouched.
- Skill + docs: browser-automation SKILL.md multi-tab/allowlist
section; docs/platform-browser.md comparison doc (new file).
- Tests (mock/fake only, no network, no real Chromium, no real
credentials): cloud.tests.test_playwright +
cloud.tests.test_websearch_worker, 48 tests OK, including 6
new wiring tests (switch off passes no variable, switch on
reaches MCP env verbatim, empty never written, old-tenant
merge idempotent + no-rewrite, ensure no-op when identical,
allowlist enforces only after wiring).
- NOT done (see PR body): no persistent profile, no credential
vault, no approval card, no CAPTCHA bypass, in-page link
navigation does not pass the allowlist, no real-Chromium live
verification, host tinyproxy filter not reloaded.
Deploy restarts kill running tasks: the worker marks them
interrupted (worker_shutdown on graceful exit, sweep on stale
reclaim), and auto_resume_interrupted used to accept goal tasks
only, so every non-goal task killed by a deploy stayed
interrupted forever unless a human retried it (the 2026-10-30
demo task died this way).
Now worker_shutdown/sweep interruptions are re-queued for
non-goal tasks too: the platform killed the task, so the
platform re-queues it. The re-queue is a fresh run, not a
breakpoint resume: goal_resume_pending is forced to 0 and any
stale resume_message is cleared, so the new run cannot be
hijacked by the goal-resume path or an old continue message;
files in the same worktree persist. All existing guards stay:
user pause, coordinator orphans, quota (still goal-only),
backoff, the 3-strike stall guard (non-goal tasks have no
checkpoints, so each platform re-queue counts as no-progress),
and the auto_resume_max cap. Other non-goal reasons (NULL,
quota, user cancel) still do not auto-resume.
- cloud/tree.py: structured worktree tree (dirs first, ignore
.git/node_modules, symlinks not followed, entry/depth
truncation flags, text render for injection)
- GET /tasks/{id}/tree with /files auth + worktree resolution,
max_depth 1..5, max_entries 1..500, escape 400, missing 404,
cloud/cross-host 409
- MCP get_task_tree translating to the new endpoint
- worker first-round injection of a repo-checkout top-level
snapshot (2 layers, 120 entries) after repo resolution,
labelled as checkout snapshot not worktree; skipped for
coordinator/resume/fork, silent when no host source
- docs/project-tree.md; tests for tree, endpoint, injection,
MCP translation
PR #145 added docker_run_* lines only to _run_in_container, but
goal/goal-resume/goal-resume-breakpoint call _call_runner directly
(to keep full stdout for goal conclusions), so the production live
test (a goal task) produced zero docker_run_ lines. Extract a
_docker_call_staged helper and route all four task paths through it;
error lines log the class name then re-raise so each caller's
kill+failed handling is unchanged.
Also fix audit attribution: audit_log has no task_id column, and
worker_exec_done/error details carried no task id, so per-task
queries missed them (the rows existed). All three exec details now
start with task_id=. Tests: 2 new goal-path stage cases + a done
detail task-id assertion (8 total in test_prestart_stages).
Claim-to-dispatch stalls (cf. 2026-10-29 task A: 61s to interrupted
with zero detail) were invisible because docker run blocks with
captured output and watchdog kills were silent. Add three layers:
- worker_exec_start/done/error audit rows in _run_attempt (PG,
survives --rm), start/done pairing shows whether an attempt died
inside the executor call;
- docker_run_start/done/error log lines in DockerExecutor (container
+ image only, never the inner command/prompt);
- stdout tail (last 2000 chars) to the worker log on failed or
interrupted results, and a log line on both watchdog kill paths.
Tests: cloud/tests/test_prestart_stages.py, 6 cases incl. a
no-prompt-in-logs assertion.
2026-10-29 live test: task A was interrupted 61s after claim but had
never written anything (no host-state entry, no worktree, no trace,
no checkpoint); auto-resume ran goal-resume, which host rightly
refused, and the task failed. Refusal only makes sense for tasks
with progress but a lost breakpoint (those go through the P0
checkpoint rebuild); for a never-started task the only meaningful
outcome is a fresh run, identical to its first attempt (same task
id, same goal, worktree does not exist yet so nothing is
overwritten).
Change (claim path only, via _resolve_breakpoint_resume):
- rebuild from checkpoint first, unchanged;
- if rebuild fails AND _task_never_started holds (state entry,
checkpoint row, worktree dir, trace file all absent), clear the
breakpoint flag so the run takes the fresh goal path, with a
worker_fresh_restart audit event;
- any evidence of prior execution, or any check failure,
keeps the old refusal path (fail-closed); the auto-resume
count cap and no-progress stall guard still bound retries.
Tests: cloud/tests/test_neverstart_fresh.py, 10 items (fallback
holds, entry/checkpoint/worktree/trace presence each block it,
corrupt state blocks it, rebuild still wins when a checkpoint
exists); test_goal_resume's breakpoint-routing test now sets up
a task with a checkpoint so it still exercises goal-resume.
Full suites: cloud 1178 (skipped 4), host 104, deploy 39, green.
run_all previously sized its pool from the ready snapshot at round start (min(capacity, ready_count)). If a round started with only 1 ready task (e.g. a just-claimed PM), only 1 thread ran; that thread entered a marker wait (up to 3600s) after its model turn, and children created/retried mid-round had no free thread to claim them. run_loop is blocked joining that round, so no new round starts; if the other worker had already scaled down, the queued child waited out the full wait timeout unclaimed. Fix: always start capacity threads. Idle threads exit immediately when no task is ready and nothing is active, and poll every 0.2s while something is active, so mid-round arrivals are claimed. Test: run_all with capacity=2, 1 initial task whose mocked run_one creates a child then polls for it; without the fix the PM poll times out in 5s (1 result), with the fix both tasks dispatch in the same round (0.2s).
Marker waits run after _run_attempt has cleared _running, so the run_loop heartbeat thread no longer refreshes the waiting parent. sweep_stale_tasks (default 300s) then kills a parent waiting up to the 3600s marker timeout, as seen in the marker live test (PM2 swept mid wait-threads). Poll and wait loops now heartbeat the parent each iteration (interval clamped to <=60s, well under the 300s threshold), best-effort, without touching _running, so no new capacity leak and cancel still works (empty _running entry has no proc/container to kill). Tests: 2 new sweep-survival tests (stale heartbeat at -299s, 2s wait, sweep must spare the parent); both fail 2/2 on pre-fix code and pass with the fix.
host