Block the round on a test-reviewer GAP for an untested Approach item #1113

New issue

Open

opened 2026-10-18 17:03:37 +00:00 by jared · 1 comment

jared commented

2026-10-18 17:03:37 +00:00

Owner

Problem

A test-reviewer GAP on an untested Approach or Required work item does not block the round.
The behavior then fails later at behavior-verifier, after several wasted rounds.

Observed

Run 64 (#1104): test-reviewers flagged that the Approach item "Record Boot and Boot::Inputs::* as the entry hop" had no test (handoffs of dispatches 568, 570, 580, 586).
The round passed anyway.
behavior-verifier then failed AC1 three times (dispatches 566, 572, 579).
The AC1 test hand-built an EntryTrace and bypassed EntryPoints, so it went green without the behavior.

Reproduce

n/a: pipeline-run behavior; evidence is run 64 in .sdlc/pipeline.db.

Origin

Trigger: run 64 dispatches 566-586
Improvised this session: none
Chain: three behavior-verifier AC1 fails ← untested Approach item passed test-reviewer ← GAP verdict is non-blocking ← DESIGN (agents/test-reviewer.md)
Where: plugins/os-sdlc/agents/test-reviewer.md
Session: de7c1b66-6981-4cf4-8801-e5e4ad89758b
Transcript: /home/jared/.claude/projects/-home-jared-dev-cc-os/de7c1b66-6981-4cf4-8801-e5e4ad89758b.jsonl

VERDICT: CREATE

## Problem A test-reviewer GAP on an untested Approach or Required work item does not block the round. The behavior then fails later at behavior-verifier, after several wasted rounds. ## Observed Run 64 (#1104): test-reviewers flagged that the Approach item "Record Boot and Boot::Inputs::* as the entry hop" had no test (handoffs of dispatches 568, 570, 580, 586). The round passed anyway. behavior-verifier then failed AC1 three times (dispatches 566, 572, 579). The AC1 test hand-built an EntryTrace and bypassed EntryPoints, so it went green without the behavior. ## Reproduce n/a: pipeline-run behavior; evidence is run 64 in .sdlc/pipeline.db. ## Origin - Trigger: run 64 dispatches 566-586 - Improvised this session: none - Chain: three behavior-verifier AC1 fails ← untested Approach item passed test-reviewer ← GAP verdict is non-blocking ← DESIGN (agents/test-reviewer.md) - Where: plugins/os-sdlc/agents/test-reviewer.md - Session: de7c1b66-6981-4cf4-8801-e5e4ad89758b - Transcript: /home/jared/.claude/projects/-home-jared-dev-cc-os/de7c1b66-6981-4cf4-8801-e5e4ad89758b.jsonl VERDICT: CREATE

jared added the

plugin/os-sdlc

filed-by/agent

category/code-change

labels

2026-10-18 17:03:37 +00:00

jared referenced this issue

2026-10-18 17:04:30 +00:00

collaborator-pass: owners_that_exist shows entry callers, not only receivers #1104

jared commented

2026-10-25 13:38:08 +00:00

Author

Owner

Occurrence: Fail test-review on a disclosed wiring gap and stop on change-auditor data-source UNKNOWNs
Seen: plugins/os-sdlc/agents/test-reviewer.md, plugins/os-sdlc/agents/change-auditor.md
Run: implementation 158 (#1312), escalation 70, session 345203e9-37ea-4caa-9b46-badf1ba9e961
Adds: test-reviewer 1368 (handoff 1395) wrote 'no test pins that wiring' and still passed; behavior-verifier failed 1363/1367/1371 on the same item until bound_exhausted:behavior-verifier/fail. Upstream cause is new: change-auditor marked both data sources UNKNOWN (handoff 1385) and the run continued to test-writer; the programmer then stubbed them with constant empty sources.

Occurrence: Fail test-review on a disclosed wiring gap and stop on change-auditor data-source UNKNOWNs Seen: plugins/os-sdlc/agents/test-reviewer.md, plugins/os-sdlc/agents/change-auditor.md Run: implementation 158 (#1312), escalation 70, session 345203e9-37ea-4caa-9b46-badf1ba9e961 Adds: test-reviewer 1368 (handoff 1395) wrote 'no test pins that wiring' and still passed; behavior-verifier failed 1363/1367/1371 on the same item until bound_exhausted:behavior-verifier/fail. Upstream cause is new: change-auditor marked both data sources UNKNOWN (handoff 1385) and the run continued to test-writer; the programmer then stubbed them with constant empty sources.

jared referenced this issue

2026-10-25 14:02:04 +00:00

Add a ticket-growth report over the pipeline DB as the baseline for pipeline realignment #1312