* fix(media): stop canonicalizeBase64 allocating one cons-string node per char
Per-character append built a rope of ~25 bytes per input character, all
live until the final join — a 20 MB base64 attachment transiently needed
~700-900 MB of heap and OOM-crashed the gateway on POST /v1/responses.
Validate in the same single pass but collect contiguous non-whitespace
runs as slices; already-canonical input is returned unchanged with zero
allocations. Measured: 15 MiB attachment 659 MB -> 0 MB transient heap,
4.3 s -> 0.2 s.
* fix(media): bound canonicalizeBase64 cleanup memory by input length
Review found the run-slicing cleanup unbounded for adversarial input:
alternating data characters and whitespace retains one slice object per
run (measured 421 MB of heap for an 8 MiB payload shredded to one run
per character). Replace the run collection with a single output buffer
materialized on the first whitespace and filled in the same validating
pass: canonical input still returns unchanged with zero allocations, and
any whitespace shape now costs at most one buffer bounded by the input
length (measured 0 MB heap delta, 210 ms for the same shredded payload).
Adds the many-short-runs regression test the review asked for, guarding
both heapUsed and arrayBuffers.
* refactor(media): condense canonicalizeBase64 invariant comment
Review asked for the repository's 1-3-line invariant form: keep why the
buffer is lazy and bounded, drop the implementation-history narration.
* chore: retrigger CI (flaky gateway-server shard)
* test(media): update base64 memory comment
Punchcard-Session: frost-cedar-willow-ae
---------
Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
The model catalog abort test waited for pid-file existence, but the
catalog child's writeFileSync exposes an open-truncate window: the
test could read the file empty and fail on a NaN pid assertion
(CI run 32231699569, checks-node-changed-extensions-config-37).
Add a qa-lab-local process-wait test helper that waits for the
asserted state (a parseable pid) with generous 10s deadlines and
self-diagnosing timeout errors, and consolidate the two hand-rolled
helper copies in model-catalog.runtime.test.ts and
test-file-scenario-command-lifecycle.test.ts onto it. Also retire
the stale 1.7s abort-settle ceiling (calibrated against a removed
1.5s in-plugin kill grace) in favor of a 5s bound with the contract
documented.
normalizeCloudRepo hand-rolled an scp-form regex that also matches
scheme URLs, so ssh://git@host:22/owner/repo put the port in the
repository path (github.com/22/owner/repo). Operators with a ported ssh
remote silently lost their cloudWorkers.projectProfiles mapping: dispatch
fell back to no-mapping behavior with no visible outcome.
Delegate to parseGitUrl, the parser project memory scope already uses for
the same host/path identity derivation, which folds userinfo, ports, and
scp-vs-scheme forms and rejects traversal. Only the config-specific
lowercase folding stays local.
* fix(line): deliver a location LINE cannot render instead of dropping it
A location whose title or address is blank makes LINE reject the whole
request, so every delivery path filtered it out before sending. The reply
then arrived without the pin, and nothing told the user or the operator that
a part of the message had been removed: the coordinates the sender supplied
were discarded silently.
The location builder now answers with the values the sender wrote — the
non-blank label plus the coordinates — as a text message, so an unrenderable
pin degrades into something the chat can show instead of disappearing. Both
delivery paths inherit that from the single builder, and the three call sites
that used to skip a null result no longer have a branch to take.
Live validation against the Messaging API confirms the shape: the authored
location is rejected with "May not be empty" on messages[0].address, the same
rejection kills an entire batch that also carries valid text, and both
degraded forms validate cleanly.
* fix(line): bound the location fallback to LINE's text limit
Nothing caps the location labels at the schema, so a long title with a blank
address produced one raw text message past LINE's 5,000-character limit — the
provider rejected it and the location was lost exactly as silently as before.
The pin path already caps each label at 100 characters. Name that limit and
apply it in the fallback too, so both forms carry the same bound instead of
one of them re-deriving the provider's rules.
* test(line): merge the duplicate send.js imports
* fix(cli): resolve infer command owners via --agent and systemAgent fallback
Multi-agent rosters (agents.ownership=explicit) dead-ended every infer
runner command: resolveDefaultAgentId threw AgentSelectionRequiredError
telling callers to pass --agent, a flag none of the commands registered,
and agents.defaults.systemAgent.agentId was ignored.
Route model run, image generate/edit/describe/describe-many, video
generate, embedding create, and model auth logout through the existing
resolveCapabilityProviderAgentId chain (explicit --agent, then
agents.defaults.systemAgent.agentId, then the sole agent) with
per-command error surfaces, and register --agent on each command.
* docs(config): scope systemAgent infer fallback help to agent-owned commands
* refactor(gateway): interrupt at admission; sessions.steer delegates to chat.send
chat.send queueMode interrupt now aborts the captured current operation
at the session-work admission owner (exact-instance abort, never a
same-key successor) and dispatches the new turn after the abort
settles; channel-originated interrupts share the same primitive.
sessions.steer keeps its validation, sharing/access checks, and
response shape but becomes a thin deprecated adapter over that path -
its private inflight/retry and hard-abort machine is deleted. The
Control UI /redirect command rides canonical chat.send. Compaction's
use of the old hard-interrupt helper moves unchanged to
session-run-interruption.ts.
* fix(gateway): interrupt non-reply admissions
Resolve and persist the exact same-origin repository that owns a live managed checkout before snapshot refs, branch deletion, or worktree removal. Fail closed on different-origin substitution. Closes#126304
* feat(ui): unify focused presentation routes
/focus/<target> replaces unshipped standalone query links across dashboard, terminal, desktop, and native apps.
Gateway-served index assets are anchored so nested documents resolve their bundles from the Control UI base path.
* test(gateway): narrow emitted asset URLs
Fixes check:test-types TS18048/TS2322 by dropping unmatched optional captures before comparing emitted asset URLs.
* test(docs): follow centralized cloud secret guidance
Fixes the stale current-main docs test after #126132 centralized GCP and Hetzner setup in docker-vm-runtime.
* test(ui): retry missing locator reads
The 500ms locator text read can time out while the menu label is still rendering, causing expect.poll to reject instead of using its owning 10s retry window. Treat only Playwright TimeoutError as a missing value so the outer poll retries while page-closure and arbitrary failures still surface.
* test(android): capture TLS probe coroutine
The TLS probe test inferred its coroutine from mutable scope children, racing unrelated child startup and teardown in CI. Capture the exact Job from inside the probe coroutine and join that owner before asserting the stale-attempt guard.
* fix(gateway): preserve plugin focus routes
Keep approval handling ahead of plugin dispatch, but treat focus documents as an unclaimed Control UI fallback after plugin authentication and routing. Exact and prefix plugin routes therefore retain ownership, while unclaimed reads serve the focus document and other methods return 404.
* fix(ui): migrate released terminal links
Preserve stable v2026.7.1 terminal query compatibility by rewriting the root/base ?view=terminal URL once to the canonical /focus/terminal path with history.replace. Keep URL parsing path-only, and leave the removed desktop and dashboard query forms as a hard cut.
* test(codex): assign run-attempt tools shard
Cached filtered configs caused duplicate ownership, and the test lacked a canonical full-suite owner.
* test(ui): keep cloud recovery proof state-owned
The recovery test should assert owner state and reload identity, while dedicated tests own transient alert visibility.
* test(qa): wait for outbound bus state
* fix(qa): reserve gateway ports through staging
* refactor(qa): keep socket creation in gateway owner
* fix(update): validate target config before handoff
Run the existing dev candidate preflight from both managed update entry points while the serving Gateway remains alive. Reject target-incompatible config before starting a handoff or scheduling restart. Closes#126269
* fix(protocol): restore gateway barrel line budget
Route SessionsPatchResult through the existing type-only schema barrel so the public export stays intact without exceeding the core max-lines gate.
* fix(ci): restore required PR gates
Keep schema-types as the registry-free wrapper, drop a redundant project schema re-export to restore the line budget, claim run-attempt-tools in one canonical shard, and align campaign tests with the new preflight boundary.
* fix(gateway): make activeRunIds presence mean a complete exact run set
Session rows no longer emit activeRunIds: [] while hasActiveRun is
true. Presence now means the complete exact set of direct run ids;
omission means identities are unavailable (projected/embedded owners);
[] only ever represents proven idle. Consumers stop guessing:
soleActiveSessionRunId() replaces the arbitrary [0] fallbacks in the
observer digest, transcript cache key, activity inspector, and
stale-terminal reconciliation, each falling back to its owner fact.
Follows the maintainer direction from #125983: the field stays as
Gateway-owned exact facts; producer-side liveness/observer projections
are a named follow-up.
* fix(gateway): clear unavailable active run ids in events
* fix(gateway): preserve idle active run sets
* fix(clients): close active run id cache gaps
* test(android): isolate history run snapshot
Use the durable run-session identity for nested session lookups on multi-agent hosts while retaining existing visibility and agent-to-agent authorization.
Co-authored-by: Ayaan Zaidi <hi@obviy.us>
* fix: capture GitHub identity from authenticated sign-in
Automatically persist verified GitHub identities from Cloudflare Access and Tailscale Serve while keeping public Git co-author credit as a separate opt-in.
* test: stabilize cleanup and activity capture
* fix(security): bind GitHub profiles by account id
* test: scope activity capture to route
* fix(security): gate profile requests on identity sync
* fix(security): close pending profile authorization gaps
* test(ui): stabilize terminal continuation menu
* test: stabilize startup recovery timing
* test: keep one Codex attempt tools owner
* fix(plugins): allow profile-independent gateway reads
alexeysophia