coilyco-flight-deck/ward

Fork 0

Add CI smoke coverage for engineer dispatch #931

New issue

Closed

opened 2026-07-10 03:41:24 +00:00 by coilyco-ops · 5 comments

coilyco-ops commented

2026-07-10 03:41:24 +00:00

Member

Context:

Engineer dispatch regresses often enough that ward needs CI coverage for the launch path, not only unit coverage around isolated helpers.

Fresh failure observed from a read-only Codex director surface on 2026-07-10:

warded engineer coilyco-flight-deck/agentic-os --instructions-file /scratch/engineer-dispatch-smoke.md could file coilyco-flight-deck/agentic-os#400 only when WARD_CONFIG_REF=forgejo.coilysiren.me/coilyco-flight-deck/agentic-os@main//.ward was set explicitly.
Without that config ref, the surface used the neutral example config and rejected coilyco-flight-deck as untrusted.
With that config ref, print mode resolved a valid launch plan, but the real run attempted local Docker from the read-only surface and failed with Cannot connect to the Docker daemon at unix:///var/run/docker.sock instead of routing through the host dispatch broker.
Related existing issues: ward#900 and ward#876.

Desired CI coverage:

Add a dispatch smoke test that runs in CI against ward itself and fails before release when warded engineer <trusted-ref> cannot resolve the real fleet config.
Add coverage for a read-only director/surface environment with WARD_READONLY=1, WARD_DISPATCH_BROKER_ADDR, and WARD_DISPATCH_BROKER_TOKEN set, asserting engineer launch goes through the broker instead of touching /var/run/docker.sock locally.
Make the test hermetic. It should not require a live LLM, a real issue mutation, or a push to main. Use an existing fake broker/test server or add one that records the requested broker operation and target.
Include the regression shape in assertions: trusted owner must include coilyco-flight-deck, and missing/unmounted Docker in the surface must not be the failure mode when a broker is configured.
Wire it into the normal CI workflow so a green PR means the engineer dispatch path still starts.

Done condition:

CI has a test that fails on the regression observed above.
The test covers both config selection and broker-vs-local-Docker routing for engineer dispatch.
Any docs or test fixtures explain why the smoke exists and link this issue plus ward#900/ward#876 where appropriate.

Context: Engineer dispatch regresses often enough that ward needs CI coverage for the launch path, not only unit coverage around isolated helpers. Fresh failure observed from a read-only Codex director surface on 2026-07-10: * `warded engineer coilyco-flight-deck/agentic-os --instructions-file /scratch/engineer-dispatch-smoke.md` could file `coilyco-flight-deck/agentic-os#400` only when `WARD_CONFIG_REF=forgejo.coilysiren.me/coilyco-flight-deck/agentic-os@main//.ward` was set explicitly. * Without that config ref, the surface used the neutral example config and rejected `coilyco-flight-deck` as untrusted. * With that config ref, print mode resolved a valid launch plan, but the real run attempted local Docker from the read-only surface and failed with `Cannot connect to the Docker daemon at unix:///var/run/docker.sock` instead of routing through the host dispatch broker. * Related existing issues: ward#900 and ward#876. Desired CI coverage: * Add a dispatch smoke test that runs in CI against ward itself and fails before release when `warded engineer <trusted-ref>` cannot resolve the real fleet config. * Add coverage for a read-only director/surface environment with `WARD_READONLY=1`, `WARD_DISPATCH_BROKER_ADDR`, and `WARD_DISPATCH_BROKER_TOKEN` set, asserting engineer launch goes through the broker instead of touching `/var/run/docker.sock` locally. * Make the test hermetic. It should not require a live LLM, a real issue mutation, or a push to main. Use an existing fake broker/test server or add one that records the requested broker operation and target. * Include the regression shape in assertions: trusted owner must include `coilyco-flight-deck`, and missing/unmounted Docker in the surface must not be the failure mode when a broker is configured. * Wire it into the normal CI workflow so a green PR means the engineer dispatch path still starts. Done condition: * CI has a test that fails on the regression observed above. * The test covers both config selection and broker-vs-local-Docker routing for engineer dispatch. * Any docs or test fixtures explain why the smoke exists and link this issue plus ward#900/ward#876 where appropriate.

coilyco-ops commented

2026-07-10 03:41:52 +00:00

Author

Member

WARD-RESERVATION: held 🔒

reservation details

Holder: container engineer-codex-ward-931 on host kais-macbook-pro-2.local.

Reserved by ward agent --harness codex (reserved 2026-07-10T03:41:51Z). Concurrent ward agent runs are blocked until it finishes or the reservation goes stale (1h TTL). --force overrides.

Do not comment on or edit this issue to steer the run while it is reserved. The engineer seeded the body once at launch and never re-reads it, so a comment or edit reaches only human readers, never the running engineer. A correction goes to a new issue, dispatched fresh. That is the only channel that reaches a run in flight. Where the forge supports it, ward locks this conversation to make that a road-block rather than a convention (ward#494).

run seed context — what this run is carrying (ward#609)

Resolved: coilyco-flight-deck/ward#931 · branch issue-931 · harness codex · workflow pull-requests
Run: engineer-codex-ward-931 · ward v0.554.0 · dispatched 2026-07-10T03:41:51Z
Comment thread: 0 included in the pre-flight read, 0 stripped (ward's own automated comments).

Static container doctrine and seed boilerplate are identical every run and omitted here (they ride ward v0.554.0).

— Codex, via ward agent

WARD-RESERVATION: held 🔒 <details><summary>reservation details</summary> Holder: container `engineer-codex-ward-931` on host `kais-macbook-pro-2.local`. Reserved by `ward agent --harness codex` (reserved 2026-07-10T03:41:51Z). Concurrent `ward agent` runs are blocked until it finishes or the reservation goes stale (1h TTL). `--force` overrides. **Do not comment on or edit this issue to steer the run while it is reserved.** The engineer seeded the body once at launch and never re-reads it, so a comment or edit reaches only human readers, never the running engineer. A correction goes to a **new issue, dispatched fresh**. That is the only channel that reaches a run in flight. Where the forge supports it, ward locks this conversation to make that a road-block rather than a convention (ward#494). <details><summary>run seed context — what this run is carrying (ward#609)</summary> - **Resolved:** `coilyco-flight-deck/ward#931` · branch `issue-931` · harness `codex` · workflow `pull-requests` - **Run:** `engineer-codex-ward-931` · ward `v0.554.0` · dispatched `2026-07-10T03:41:51Z` - **Comment thread:** 0 included in the pre-flight read, 0 stripped (ward's own automated comments). Static container doctrine and seed boilerplate are identical every run and omitted here (they ride ward v0.554.0). </details> </details>  — Codex, via `ward agent`

coilyco-ops commented

2026-07-10 03:52:23 +00:00

Author

Member

WARD-RESERVATION: released 🛑

release details

Run never started. ward container reap released container ward (--harness codex): it exited without launching the agent (smoke-test death, ward#222/#264/#595), so it did no work and the hold it took is retracted. Nothing is running on this issue. It needs re-dispatch. A ward agent director re-queues it automatically. A manual ward agent retry no longer needs --force.

— Codex, via ward agent

WARD-RESERVATION: released 🛑 <details><summary>release details</summary> Run never started. `ward container reap` released container `ward` (`--harness codex`): it exited without launching the agent (smoke-test death, ward#222/#264/#595), so it did no work and the hold it took is retracted. Nothing is running on this issue. It needs re-dispatch. A `ward agent director` re-queues it automatically. A manual `ward agent` retry no longer needs `--force`. </details>  — Codex, via `ward agent`

coilyco-ops commented

2026-07-10 03:53:41 +00:00

Author

Member

WARD-RESERVATION: released 🛑

release details

— Codex, via ward agent

coilyco-ops commented

2026-07-10 03:54:26 +00:00

Author

Member

WARD-RESERVATION: released 🛑

release details

— Codex, via ward agent

coilyco-ops referenced this issue from a pull request that will close it,

2026-07-10 03:56:06 +00:00

Add CI smoke coverage for engineer dispatch #933

coilyco-ops commented

2026-07-10 03:57:02 +00:00

Author

Member

WARD-OUTCOME: submitted

details

workflow: pull-requests; review summary: skipped by ~/.ward/config.yaml default
retrospective: the smoke was straightforward once the brokered freeform path was isolated
confidence: high
surprises: none
follow-ups: none

WARD-OUTCOME: submitted <details><summary>details</summary> workflow: pull-requests; review summary: skipped by ~/.ward/config.yaml default retrospective: the smoke was straightforward once the brokered freeform path was isolated confidence: high surprises: none follow-ups: none </details>