/acceptance

/wicked-testing:acceptance

Evidence-gated 3-agent acceptance testing pipeline. Writer designs test plans, Executor collects artifacts, Reviewer evaluates independently — eliminating self-grading false positives.

Usage

/wicked-testing:acceptance <scenario-file> [--phase write|execute|review|all] [--json]

• scenario-file — path to a wicked-testing scenario .md file (required)
• --phase all (default) — full Write → Execute → Review pipeline
• --phase write — generate test plan only
• --phase execute — run execution with existing plan
• --phase review — review existing evidence
• --json — emit JSON envelope

Instructions

1. Validate Input

Check scenario file exists:

test -f "{scenario-file}" || echo "ERR_SCENARIO_NOT_FOUND"

Check config:

test -f ".wicked-testing/config.json" || echo "ERR_NO_CONFIG"

2. Set Up Run Context

Create the run record FIRST so the evidence directory is named with the run's canonical UUID. This avoids the 1-second-granularity timestamp collision that used to stack two concurrent runs into the same dir, and it matches the public-contract path the schemas document.

// After ensuring project + scenario exist in DomainStore:
const run = store.create('runs', { project_id, scenario_id, started_at, status: 'running' });
const EVIDENCE_DIR = `.wicked-testing/evidence/${run.id}`;
// mkdir -p EVIDENCE_DIR; then:
store.update('runs', run.id, { evidence_path: EVIDENCE_DIR });

3. Invoke the acceptance-testing skill

The acceptance-testing skill (see skills/acceptance-testing/SKILL.md) orchestrates the 3-agent pipeline.

Follow the instructions in that skill file exactly, including:

• Dispatching acceptance-test-writer for the test plan
• Dispatching acceptance-test-executor with test plan + evidence directory
• Dispatching acceptance-test-reviewer with ONLY evidence paths (REVIEWER ISOLATION)
• Writing verdict and updating run in DomainStore

4. REVIEWER ISOLATION (CRITICAL)

The reviewer MUST receive only:

• Scenario file path
• Evidence directory path
• Test plan (written to evidence directory by executor)

The reviewer MUST NOT receive:

• Executor stdout/stderr directly
• Executor reasoning or explanations
• Any executor conversational context

See skills/acceptance-testing/SKILL.md for the exact dispatch prompt.

5. Output

Without --json — Return the verdict with acceptance criteria table.

With --json — Emit the JSON envelope:

python3 -c "import json,sys; sys.stdout.write(json.dumps({'ok': True, 'data': {'verdict': 'PASS', 'run_id': '...', 'evidence_path': '...', 'assertions_passed': N, 'assertions_failed': 0}, 'meta': {'command': 'wicked-testing:acceptance', 'duration_ms': 0, 'schema_version': 1, 'store_mode': '...'}}))" 2>/dev/null || python -c "..."

On FAIL, ok remains true but data.verdict is 'FAIL' and data.failures lists the failures.