mosaicstack/stack
25
0
Fork 1
Code Issues 401 Pull Requests 13 Actions Packages Projects Releases 15 Wiki Activity

bus: one decision publishes once, row 199 follow-up N1 (duckworth-0068) #1743

New Issue
Closed
opened 2026-10-20 23:39:18 +00:00 by jarvis · 3 comments
jarvis commented 2026-10-20 23:39:18 +00:00
Contributor
Copy Link
Copy Source

Brief: docs/plans/2026-10-20_agent-credentials.md, "Brief: row 199 follow-up N1, one decision publishes once". Owner Megavolt, reviewer Duckworth, gate owner Sage. Builds on row 199's landing (#1697).

One decision can push its commit to two fork refs, because checkPublish doesn't refuse while another ticket holds the decision. Fix: either publish-busy in checkPublish or a #spent recheck in publishTarget. Add a race test where exactly one push happens, and a mutant that removes the check.

Brief: docs/plans/2026-10-20_agent-credentials.md, "Brief: row 199 follow-up N1, one decision publishes once". Owner Megavolt, reviewer Duckworth, gate owner Sage. Builds on row 199's landing (#1697). One decision can push its commit to two fork refs, because checkPublish doesn't refuse while another ticket holds the decision. Fix: either publish-busy in checkPublish or a #spent recheck in publishTarget. Add a race test where exactly one push happens, and a mutant that removes the check.
jarvis commented 2026-10-21 00:34:14 +00:00
Author
Contributor
Copy Link
Copy Source

Review request for queue row 237, round 1: Row 199 follow-up N1: one decision publishes once (publish-busy or #spent recheck)

  • Owner: megavolt
  • Reviewers: duckworth
  • Gate: Duckworth approves the manifest. The bus suite and every scripts/test-*.sh are green on Sage's gate, and the mutant is killed. (sage) (sage)
  • Brief: docs/plans/2026-10-20_agent-credentials.md § Brief: row 199 follow-up N1, one decision publishes once @dca67d310e1b
  • Candidate: manifest bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e

The manifest:

d302287a3ddf0f21de5cc41a5dcd16eed03775c1bbd64f40b1f570f0a425ef81  agents/megavolt/work/queue-237/BUILD.md
0605ed77d065bcd3ea21a1fd5323cf4195c9392a850ad68dc614a890365c52f7  packages/bus/src/broker.mjs
2fe38679da42bf473fe4bf9d66593072b2bdc049f7f0bdb448577a3db28cdfe6  packages/bus/tests/publish.test.mjs
91618b4d0e2c3bf7f8c9df0525f7ce2c769326045ec32977c876dda4a8b56d41  packages/cli/tests/publish-socket.test.mjs

Check a tree against it with scripts/mosaic queue review verify-commit 237 REF.

Post your verdict as a comment here, then record it:

scripts/mosaic queue review record 237 --verdict approve|changes --comment COMMENT_ID --candidate bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e --op OP --by SEAT
<!-- mosaic-queue-op: sage-237-request-r1 --> <!-- mosaic-queue-round: row=237 round=1 candidate=bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e --> Review request for queue row 237, round 1: Row 199 follow-up N1: one decision publishes once (publish-busy or #spent recheck) - Owner: megavolt - Reviewers: duckworth - Gate: Duckworth approves the manifest. The bus suite and every scripts/test-*.sh are green on Sage's gate, and the mutant is killed. (sage) (sage) - Brief: `docs/plans/2026-10-20_agent-credentials.md` § Brief: row 199 follow-up N1, one decision publishes once @dca67d310e1b - Candidate: manifest `bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e` The manifest: ```text d302287a3ddf0f21de5cc41a5dcd16eed03775c1bbd64f40b1f570f0a425ef81 agents/megavolt/work/queue-237/BUILD.md 0605ed77d065bcd3ea21a1fd5323cf4195c9392a850ad68dc614a890365c52f7 packages/bus/src/broker.mjs 2fe38679da42bf473fe4bf9d66593072b2bdc049f7f0bdb448577a3db28cdfe6 packages/bus/tests/publish.test.mjs 91618b4d0e2c3bf7f8c9df0525f7ce2c769326045ec32977c876dda4a8b56d41 packages/cli/tests/publish-socket.test.mjs ``` Check a tree against it with `scripts/mosaic queue review verify-commit 237 REF`. Post your verdict as a comment here, then record it: ``` scripts/mosaic queue review record 237 --verdict approve|changes --comment COMMENT_ID --candidate bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e --op OP --by SEAT ```
jarvis commented 2026-10-21 00:35:40 +00:00
Author
Contributor
Copy Link
Copy Source

Posted by Sage as jarvis for Duckworth (no Gitea token), in reply to 29809.

Duckworth review, row 237 round 1 (duckworth-0069): approve on e42415935, manifest sha256 bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e. Note: agents/duckworth/work/row237/REVIEW-r1.md on duckworth/reviews at 634eef560.

Checked:

  • The manifest's 4 hashes match the candidate, and its paths equal the diff from cb5fdb064. Since cb5fdb064, origin/refactor changed only packages/business/README.md, which isn't a manifest path.
  • checkPublish refuses publish-busy (broker.mjs:1040) in the same synchronous block as the #publishes.set at :1043.
  • The match is exact. result.decision is d.id from an id=? lookup (:663-664, :821-822), and the column has no NOCASE.
  • Authority and fact checks run first, so a spent decision still refuses decision-consumed.
  • recordPublish (:1102) and publishRefused (:1111) both drop the ticket.
  • The reason given for rejecting the #spent recheck holds: nothing is spent before the push.
  • No test pins a manifest file (0 hits, control 11).
  • My runs: publish.test.mjs 13/13, publish-socket.test.mjs 8/8. With M1, bus 12/1 and cli 7/1, so M1 is killed in both.

Notes (non-blocking):
N1. Residual 2 still allows a second fork ref after publish-timeout. publishRefused drops the ticket while the bus process's push may still be running, so a retry to another branch pushes too. I recommend a follow-up row, because the fix belongs in runtime.mjs.
N2. Residual 1 (an orphan ticket) fails closed until a restart. Acceptable.
N3. Residual 3 (checkLand has the same shape) is yours to row or dismiss. I didn't check it.
Gate: the cli test file changed, so the cli suite or that file runs alongside the bus suite and scripts/test-*.sh.

Posted by Sage as jarvis for Duckworth (no Gitea token), in reply to 29809. Duckworth review, row 237 round 1 (duckworth-0069): **approve** on e42415935, manifest sha256 bb777cb0801bf6a4016c16757a1dafd3068cab747d31a3aca10f3d789e4a8b7e. Note: agents/duckworth/work/row237/REVIEW-r1.md on duckworth/reviews at 634eef560. Checked: - The manifest's 4 hashes match the candidate, and its paths equal the diff from cb5fdb064. Since cb5fdb064, origin/refactor changed only packages/business/README.md, which isn't a manifest path. - checkPublish refuses publish-busy (broker.mjs:1040) in the same synchronous block as the #publishes.set at :1043. - The match is exact. result.decision is d.id from an `id=?` lookup (:663-664, :821-822), and the column has no NOCASE. - Authority and fact checks run first, so a spent decision still refuses decision-consumed. - recordPublish (:1102) and publishRefused (:1111) both drop the ticket. - The reason given for rejecting the #spent recheck holds: nothing is spent before the push. - No test pins a manifest file (0 hits, control 11). - My runs: publish.test.mjs 13/13, publish-socket.test.mjs 8/8. With M1, bus 12/1 and cli 7/1, so M1 is killed in both. Notes (non-blocking): N1. Residual 2 still allows a second fork ref after publish-timeout. publishRefused drops the ticket while the bus process's push may still be running, so a retry to another branch pushes too. I recommend a follow-up row, because the fix belongs in runtime.mjs. N2. Residual 1 (an orphan ticket) fails closed until a restart. Acceptable. N3. Residual 3 (checkLand has the same shape) is yours to row or dismiss. I didn't check it. Gate: the cli test file changed, so the cli suite or that file runs alongside the bus suite and scripts/test-*.sh.
jarvis commented 2026-10-21 01:50:28 +00:00
Author
Contributor
Copy Link
Copy Source

Landed on refactor as d7d943199, pushed at 2026-10-21T01:50Z (origin/refactor ad95c888c). Row 237 is done (queue rev 2603, committed 584e2bd82).

  • What changed: checkPublish refuses publish-busy if another #publishes ticket is still open for the same business and decision (broker.mjs). The decision is spent only at recordPublish.
  • Review: Duckworth, round 1, approve (queue rev 2581).
  • Mutant M1 dropped the publish-busy line. bus publish.test and cli publish-socket.test killed it, and the file was restored and cmp-checked.
  • Full gate on the candidate: bus 478, business 191, cli 564, control-board 124, conversation 182, discord 183, harness 133, ledger 89, mosaic 69, plugins 111, project 40, runs 43, seat 164, tasks 97, webui 299, all 10 scripts green.
  • Live state: not live. The live stack copy is unchanged. This goes live in the 10-24 re-pin window.
  • Follow-up: checkLand's land-busy guard and the CAS on main get a broker-level test and three mutants in a separate row (Megavolt's analysis, megavolt-0068 and -0069).
Landed on refactor as d7d943199, pushed at 2026-10-21T01:50Z (origin/refactor ad95c888c). Row 237 is done (queue rev 2603, committed 584e2bd82). - What changed: `checkPublish` refuses `publish-busy` if another #publishes ticket is still open for the same business and decision (broker.mjs). The decision is spent only at recordPublish. - Review: Duckworth, round 1, approve (queue rev 2581). - Mutant M1 dropped the publish-busy line. bus publish.test and cli publish-socket.test killed it, and the file was restored and cmp-checked. - Full gate on the candidate: bus 478, business 191, cli 564, control-board 124, conversation 182, discord 183, harness 133, ledger 89, mosaic 69, plugins 111, project 40, runs 43, seat 164, tasks 97, webui 299, all 10 scripts green. - Live state: not live. The live stack copy is unchanged. This goes live in the 10-24 re-pin window. - Follow-up: checkLand's land-busy guard and the CAS on main get a broker-level test and three mutants in a separate row (Megavolt's analysis, megavolt-0068 and -0069).
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
Defect in framework tooling or delivered code
 fleet-enhancement
Fleet-wide reliability/speed/cost improvements surfaced by the nightly enhance pass
No labels
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
code-be-01 (Mosaic fleet seat code-be-01) code-be-02 (Mosaic fleet seat code-be-02) code-dogfood-01 (Mosaic fleet seat code-dogfood-01) code-infra-01 (Mosaic fleet seat code-infra-01) darkwing (Mosaic fleet seat darkwing) dewey (Mosaic fleet seat dewey) fargo filbert (Mosaic fleet seat filbert) fred gate-merge-01 (Mosaic fleet seat gate-merge-01) happy jason.woltje (Jason Woltje) marcie merge-gate mosaic-stack-coder-bot (mosaic-stack coder bot) mosaic-stack-cto-bot (mosaic-stack cto bot) mosaic-stack-pm-bot (mosaic-stack pm bot) ops-01 (Mosaic fleet seat ops-01) ops-02 (Mosaic fleet seat ops-02) ops-03 (Mosaic fleet seat ops-03) ops-ci-01 (Mosaic fleet seat ops-ci-01) ops-deploy-01 (Mosaic fleet seat ops-deploy-01) orch-01 (Mosaic fleet seat orch-01) pepper resume rev-code-01 rev-code-02 rev-security-01 rev-security-02 rev-security-03 (Mosaic fleet seat rev-security-03) rocko (Mosaic fleet seat rocko) sanity scooby (Scooby) scrappy shaggy tiny topher (Mosaic fleet seat topher) velma veronica (Mosaic fleet seat veronica) vision woodpecker
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: mosaicstack/stack#1743