lidge-jun/codexclaw / plugins/codexclaw/skills/qa
cxc-qa
VerifiedUse for manual web, TUI, CLI, API and desktop QA. Triggers: smoke test, visual QA, screenshot check, CJK clipping, 수동 QA, 실제로 되는지 확인, 동작 확인, 직접 돌려봐.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.1 KB |
| agents/openai.yaml | config | 155 B |
| references/adversarial-oracles.md | markdown | 2.5 KB |
| references/cli-tui-qa.md | markdown | 3.2 KB |
| references/evidence-contract.md | markdown | 3.7 KB |
| references/http-api-qa.md | markdown | 2.9 KB |
| references/scope-and-binding.md | markdown | 1.3 KB |
| references/teardown-receipts.md | markdown | 392 B |
| references/visual-qa.md | markdown | 7.1 KB |
| scripts/validate-evidence.mjs | script | 21.1 KB |
references/http-api-qa.md
View on GitHub# HTTP API QA — wire-driving scenarios (QA-HTTP-01)
Drive the real wire and capture what it actually returned. E7 discipline.
Ownership: this file owns the DRIVE-the-surface procedure only — test
strategy/harness choice is `dev-testing` (references/backend-testing.md);
the response-envelope DEFINITION is `dev-backend` §5. Assert against the
repo's envelope; never restate its rules here.
## Faithful channel
- `curl -i` (or `-v` when TLS/redirect behavior is the claim) — headers AND
body in one artifact. Client SDK output is not wire evidence: it may
deserialize, retry, or normalize before you see it.
- One scenario = one captured request/response pair. Multi-step flows
(login -> use token) capture every hop, numbered in `invocation.txt`.
- Streaming/SSE exception: `curl -N --max-time <n>` with the first N events
captured; state the cutoff. WebSocket flows need a scripted client — name
the script in the artifact.
## Scenario axes (probe the applicable ones)
1. **Auth states** — anonymous, valid token, expired token, wrong
scope/role. Each returns the DOCUMENTED status (401 vs 403 distinction is
a finding when collapsed); no protected payload ever leaks in an error
body.
2. **Contract shape** — the repo's envelope (see `dev-backend` §5) holds on
BOTH success and error paths; error responses carry the machine `code`
(clients must never need to parse `message`). Capture one success + one
error body per surface as the shape evidence.
3. **Idempotency / repeat** — double-submit the same POST/PUT; assert
documented behavior (dedupe, 409, or duplicate side effect = finding).
Retried requests honor idempotency keys when the API claims them.
4. **Boundary payloads** — empty body, missing required field, oversized
payload (documented limit +1), wrong `Content-Type`. Expect a 4xx with a
parseable error, never a 500 or a hang.
5. **Content negotiation** — wrong/absent `Accept` returns a sane default or
406, not garbage; compressed responses (`Accept-Encoding`) still parse.
6. **CORS preflight** (browser-facing APIs only) — `curl -i -X OPTIONS` with
`Origin` + `Access-Control-Request-Method`; captured allow-list matches
intent (a `*` on a credentialed API is a finding).
7. **Status-code truth table** — for the changed routes, one captured line
per (method, auth-state) cell you claim; do not infer untested cells.
## Artifacts + teardown
Per SKILL.md §3: `invocation.txt` (exact curl lines), `capture.txt`
(response(s) verbatim), `verdict.json`. Server you started for the pass ->
teardown line with `lsof -i :<port>` empty proof (SKILL.md §6). Redact
tokens/cookies in artifacts: keep length + first 4 chars only.
## Adversarial mapping
Axes 3-6 ARE SKILL.md §4 classes 1-4 specialized for the wire; class 5
(viewport/CJK) is N/A for headless APIs — record the N/A reason once per
matrix, not per scenario.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |