Menu
Por padrão, está selecionado o prompt que primeiro revisa a origem. Você pode mudar para um comando direto ou baixar uma cópia local.
Leia o SKILL.md e os arquivos complementares exibidos pelo SkillsMP antes de decidir se vai instalar.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Um comando direto ignora o prompt de revisão. Verifique a origem antes de executá-lo.
npx skills add https://github.com/T-rav/hydraflow --skill hf-audit-integration-testsO comando permanece em uma só linha. Role horizontalmente para revisá-lo antes de copiar.
Prefere uma cópia local? Baixe os arquivos disponíveis atualmente no SkillsMP.
Baseado na classificação ocupacional SOC
Exibindo SKILL.md
| name | hf.audit-integration-tests |
| description | Integration Test Audit |
Run a comprehensive integration test audit across the entire repo. Dynamically discovers external dependencies in source code, inventories existing integration tests, identifies coverage gaps, flags ugly/outdated tests, and creates GitHub issues for findings.
Resolve configuration before doing anything else:
echo "$HYDRAFLOW_GITHUB_REPO" — if set, use it as the target repo (e.g., owner/repo). If empty, run git remote get-url origin and extract the owner/repo slug (strip https://github.com/ prefix and .git suffix).echo "$HYDRAFLOW_GITHUB_ASSIGNEE" — if set, use it as the issue assignee. If empty, extract the owner from the repo slug (the part before /).hydraflow-plan as the label for created issues.$REPO, $ASSIGNEE, $LABEL.Discover project structure:
*.py source files, excluding .venv/, venv/, __pycache__/, node_modules/, dist/, build/.tests/ or test/ directory, or matching test_*.py).Launch agents in parallel using Task with run_in_background: true and subagent_type: "general-purpose":
Wait for all agents to complete.
After all finish, run gh issue list --repo $REPO --label $LABEL --state open --search "integration test" --limit 200 to show the user a final summary of all issues created.
You are an integration test auditor for the project at {repo_root}.
## Configuration
- GitHub repo: {REPO}
- Assignee: {ASSIGNEE}
- Label: {LABEL}
## Steps
### Phase 1: Map External Dependencies
1. Use Glob to find all *.py source files (exclude tests/, .venv/, __pycache__/)
2. Read each source file and catalog every external dependency:
- **Subprocess calls**: asyncio.create_subprocess_exec, subprocess.run, os.system
- **CLI tools**: git, gh, claude, docker, npm, make, or any other CLI invocations
- **HTTP clients**: httpx, requests, aiohttp, urllib calls to external services
- **Database**: SQLAlchemy, pymysql, psycopg2, sqlite3, Redis, Qdrant
- **File I/O**: Reading/writing state files, config files, log files
- **Message queues**: Celery, RabbitMQ, Kafka
- **External APIs**: Any API client calls (OpenAI, Slack, GitHub API, etc.)
For each dependency, note: file path, line number, method name, what it calls, error handling
### Phase 2: Inventory Existing Integration Tests
3. Read all test files. Catalog every test marked @pytest.mark.integration or that tests real external interactions
4. For each existing test, note: what external dependency it exercises, whether it uses real services or mocks
### Phase 3: Gap Analysis
5. Cross-reference Phase 1 dependencies against Phase 2 test coverage
6. Identify external interaction paths with ZERO integration test coverage
7. Identify tests that claim to be "integration" but are fully mocked
### Phase 4: Create GitHub Issues
8. For each finding, check for duplicate GH issues first:
gh issue list --repo {REPO} --label {LABEL} --state open --search "<key terms>"
9. Create GH issues for NEW findings only, grouped by theme:
gh issue create --repo {REPO} --assignee {ASSIGNEE} --label {LABEL} --title "Integration Test: <theme>" --body "<details>"
## Issue Body Format
```markdown
## Context
<1-2 sentences on why this integration test matters>
## External Dependencies Not Covered
| Dependency | File:Line | Method | What It Calls |
|------------|-----------|--------|---------------|
| <name> | <path:line> | <method> | <endpoint/operation> |
## Suggested Test Scenarios
- [ ] <scenario 1>
- [ ] <scenario 2>
## Notes
- Priority: <high/medium/low based on blast radius>
Create ONE issue per theme, not one per missing test. Good themes:
Be pragmatic: focus on high-blast-radius gaps. External API tests are expensive — note them but mark as low priority.
Return a summary of all findings grouped by category, with GH issue URLs created.
## Agent 2: Test Quality & Anti-Pattern Detection
You are a test quality auditor focused on integration test anti-patterns for the project at {repo_root}.
Flag these patterns:
with patch(...) blocksBe pragmatic — only flag patterns that actually hurt test reliability or maintainability.
Return a summary of all findings grouped by category, with GH issue URLs created.
## Agent 3: Cross-Module Integration Gaps
You are an integration test auditor focused on cross-module interactions for the project at {repo_root}.
make test-integration or equivalent targetFocus on the highest-value integration gaps — where module boundaries are most likely to break.
Return a summary of all findings grouped by category, with GH issue URLs created.
## Important Notes
- Each agent should read files directly (no spawning sub-agents)
- Each agent should check `gh issue list` before creating any issue to avoid duplicates
- All issues should use the resolved `$REPO`, `$ASSIGNEE`, and `$LABEL`
- Group related findings into single themed issues — don't create one issue per missing test
- Title format: "Integration Test: <theme>" for consistency
- For confirmed bugs found during audit, use title format: "Bug: <description>"
- Be pragmatic: external API tests are expensive — note them but mark as low priority. Focus on subprocess calls, file I/O, and inter-module interactions as high priority.