Menü
Codex-primary single-round test coverage review
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Conduct a decision-tree interview with the user to lock in requirements, motivation, scope, and non-goals before planning.
Stage 3 of three-stage planning pipeline. Produce a file-level implementation plan via detail-planner/detail-reviewer loop, then get user approval. Inputs are confirmed intent (<session-id>-intent.md) and outline (<session-id>-outline.md) from prior stages.
Edit source code for the current task. Delegates editing and lint/typecheck/self-repair to a subagent.
Plan and write test cases with high reasoning effort. Test iteration runs in a subagent to minimize confirmations.
Explore the codebase to understand existing patterns, constraints, and relevant files before planning.
Investigate git history, docs/history.md, and GitHub issue/PR timeline since the relevant issue opened, to surface changes that may invalidate the issue's premises.
| name | review-tests |
| description | Codex-primary single-round test coverage review |
| model | sonnet |
| effort | low |
| context | fork |
Review test case completeness against source code via Codex (single round, no re-loop).
RT-1. Identify test file(s) and source file(s). Ask the user if ambiguous.
RT-2. Assemble review input — concatenate test file(s) and source file(s) contents into <PLANS_DIR>/<session-id>-test-review.md via Write. Resolve <PLANS_DIR> via skills/_shared/resolve-plans-dir.md. Initialize EXTENSIONS_USED=0.
RT-3. Invoke "$AGENTS_CONFIG_DIR/skills/review-tests/scripts/run-codex-review-loop.sh" (Bash), exporting AGENTS_CONFIG_DIR, SESSION_ID, PLANS_DIR, EXTENSIONS_USED. The wrapper auto-adds --context test-design.md. Exit-code handling (SSOT: skills/_shared/codex-review-loop.md; exit 1 is TERMINAL):
<PLANS_DIR>/<session-id>-test-review-codex-round-1-raw.md; present gaps; suggest specific test cases → RT-4 WARNINGS (no re-loop).review-loop-summarize-concerns; present summary → RT-4 WARNINGS.test-reviewer subagent; APPROVED → RT-4 COMPLETE; NEEDS_REVISION → RT-4 WARNINGS.TOKEN=$(node "$AGENTS_CONFIG_DIR/bin/compute-staged-tests-token.js")echo "<<WORKFLOW_REVIEW_TESTS_COMPLETE: token=${TOKEN}>>"echo "<<WORKFLOW_REVIEW_TESTS_WARNINGS: token=${TOKEN} warnings=N — blocking: /write-code stays blocked until the gaps are addressed and /review-tests is re-run>>"WORKFLOW_WRITE_TESTS_NOT_NEEDED was emitted (propagated skip).The Test Case Categories checklist lives in skills/_shared/test-design.md — do not duplicate it here.
WARNINGS is BLOCKING: hooks/workflow-gate/review-tests-checker.js blocks /write-code while warnings_summary is recorded.
Emit exactly one sentinel per run: COMPLETE on pass, WARNINGS on any gap or warning.
On exit 4, emit neither sentinel and HALT.
Invariant: RT-4 emits exactly one of COMPLETE/WARNINGS; never both, never zero (except exit 4).