spectre-test

test

Add risk-weighted behavioral tests and commit standalone batches. Test boundaries, prioritize where breakage hurts, skip code that cannot break; risk assessment is inline and no intermediate report files exist.

Inputs

• $ARGUMENTS — optional explicit feature name/root or descendant artifact, scope hint, or specific files to focus on, plus --orchestrated when a parent workflow owns the next step.
• Optional orchestrator-provided risk plan from spectre-clean or spectre-ship: files already tiered P0-P3 plus batch assignment. When present, use it as the plan for this batch.

Working Set (late-bound — read at run-time, never inline)

• Reuse a managed FEATURE_ROOT only when explicit/current-thread evidence ties it to this work (physical directory wins; never branch/recency/lifecycle/scans); distinct work ignores ambient roots. Otherwise, including on collision, standalone MUST first load and follow Skill(spectre-feature-root) through DONE; orchestrated calls escalate. Keep writes beneath it and pass it unchanged.
• Repair stale feature/root metadata in artifacts this workflow touches.
• Full Working Set = UNION of: committed changes (validate any provided commit_id; invalid → STOP and ask), staged (git diff --cached --name-only), unstaged (git diff --name-only), untracked (git ls-files --others --exclude-standard). Keep the projection local/in-thread; write no working-set artifact.
• Standalone: baseline-lint the set and map import/dependency edges. All paths absolute from repo root.

Method / guardrails

• Orchestrated Ship/Clean mode: consume the supplied unchanged set/risk plan; do not redo broad analysis. Edit tests/fixtures only—never production/source; do not stage or commit. The test lead may batch testers internally, runs only new/changed focused tests, and returns compact changed-path/check results plus cross-boundary needs to the parent.
• Triage every changed file into a risk tier (inline):
  • P0 Critical — auth/payment/security/crypto/session/token, PII/permissions/user-data mutation, external handlers, DB migrations, or @critical. Cover every user-facing outcome/error path, null/empty/malformed/overflow security inputs, public API/schema, and mutation-resistant assertions.
  • P1 Core — feature components, API/state/business logic, fetch/cache. Cover public happy/error paths and exported-boundary contracts; skip internal helpers/exhaustive branches.
  • P2 Supporting — real-logic utils/validators/transformers/hooks/adapters. Cover exported happy paths; skip private/trivial functions.
  • P3 Skip — types, config, styles/docs, logic-free constants/enums/barrels/pass-throughs, generated/build tooling. Types + lint suffice; mark SKIP — {reason}.
• Write or consume the in-thread test plan (3–7 bullets, - [P{tier}] {file}: {behavior}): P0 → multiple bullets (behaviors + error paths); P1 → 1–2; P2 → 1; P3 → SKIP line.
• Dispatch @spectre_tester in parallel: one message/multiple tasks; P0 = one agent/file, P1 = 2–3 files/agent, P2 = 3–5, up to 8 agents. Give batch paths/tier and require behavioral, outcome-not-call, mutation-resistant tests; wait before verifying.
• Quality: one behavior/test; descriptive when_[cond]_then_[outcome] names; outcome assertions (calls only for prevented side effects), refactor/mutation resilience. Do not mock internals, duplicate type/framework coverage; test API/event schemas at boundaries.
• Verify before commit: standalone runs affected lint plus new/changed and related tests across demonstrated dependencies, then spot-checks quality. Branch-caused → repair/reverify; unrelated → route/continue; indeterminate → reproduce only the failing check at base. Never run a repository-wide baseline or full suite from this skill.
• Commit guard: --no-verify, eslint-disable, and committing code carrying eslint-disable are expressly forbidden without the user's explicit permission.