메뉴
Claude Code AI-assisted development workflow. Activate when discussing Claude Code usage, AI-assisted coding, prompting strategies, or Claude Code-specific patterns.
Activate when user needs multi-URL scraping, browser automation pipelines, or efficient tool orchestration to reduce API round-trips and context usage.
Language-agnostic API design patterns covering REST and GraphQL, including resource naming, HTTP methods, status codes, versioning, pagination, filtering, authentication, error handling, and schema design. Activate when working with APIs, REST endpoints, GraphQL schemas, API documentation, OpenAPI/Swagger, JWT, OAuth2, endpoint design, API versioning, rate limiting, or GraphQL resolvers.
Git workflow and commit guidelines. Trigger keywords: git, commit, push, .git, version control. MUST be activated before ANY git commit, push, or version control operation. Includes security scanning for secrets (API keys, tokens, .env files), commit message formatting with HEREDOC, logical commit grouping (docs, test, feat, fix, refactor, chore, build, deps), push behavior rules, safety rules for hooks and force pushes, and CRITICAL safeguards for destructive operations (filter-branch, gc --prune, reset --hard). Activate when user requests committing changes, pushing code, creating commits, rewriting history, or performing any git operations including analyzing uncommitted changes.
Testing workflow patterns and quality standards. Activate when working with tests, test files, test directories, code quality tools, coverage reports, or testing tasks. Includes zero-warnings policy, targeted testing during development, mocking patterns, and best practices across languages.
Ansible automation workflow guidelines. Activate when working with Ansible playbooks, ansible-playbook, inventory files (.yml, .ini), or Ansible-specific patterns.
Guidelines for containerized projects using Docker, Dockerfile, docker-compose, container, and containerization. Covers multi-stage builds, security, signal handling, entrypoint scripts, and deployment workflows.
| name | claude-code-workflow |
| description | Claude Code AI-assisted development workflow. Activate when discussing Claude Code usage, AI-assisted coding, prompting strategies, or Claude Code-specific patterns. |
| location | user |
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.
Guidelines for effective AI-assisted development with Claude Code.
| Value | Task Type | Example |
|---|---|---|
| High | Boilerplate generation | CRUD endpoints, test scaffolding |
| High | Refactoring | Extract function, rename across files |
| High | Documentation | Docstrings, README updates |
| Medium | Bug investigation | Error analysis, log interpretation |
| Medium | Code review | Pattern suggestions, security review |
| Low | Novel architecture | Requires deep domain knowledge |
| Low | Ambiguous requirements | Needs human clarification |
The RECOMMENDED workflow for most development tasks:
Explore: Use Glob, Grep, Read to understand existing code
Plan: Outline changes before implementation
Code: Implement changes incrementally
Commit: Create logical, atomic commits
1. Write failing test → 2. Implement minimum to pass → 3. Refactor → 4. Repeat
For uncertain implementations: Create throwaway spike, validate, delete, implement properly with tests.
MUST: Provide specific file paths, include error messages verbatim, state expected vs actual behavior
SHOULD: Reference previous context, specify output format, include constraints upfront
SHOULD NOT: Repeat information in context, include unnecessary background, ask multiple unrelated questions
[Initial Context] → [Specific Task] → [Iterative Refinement]
| Scenario | Recommendation |
|---|---|
| Search 5+ files for pattern | Subagent |
| Edit single file | Direct |
| Run tests + fix failures | Subagent |
| Explore unfamiliar codebase | Subagent |
| Implement planned changes | Direct |
SHOULD use for: Architecture decisions, complex debugging, multi-file refactoring, novel problem solving, security-sensitive code review
RECOMMENDED for: Routine code changes, test writing, documentation, code review, bug fixes with clear scope
MAY use for: Simple queries, syntax checking, format conversions, quick lookups, commit message generation
IF deep reasoning OR high-stakes: USE Opus
ELSE IF routine development: USE Sonnet
ELSE IF simple/quick: USE Haiku
MUST: Test execution after changes, linting/formatting, type checking, security scanning
SHOULD: Coverage thresholds, performance benchmarks, integration tests
REQUIRED: Security-sensitive changes, production deployments, architecture modifications, public API changes
RECOMMENDED: Complex business logic, performance-critical code, external integrations
Effective: "Refactor UserService in src/services/user.py to use dependency injection for database connection"
Ineffective: "Clean up the user code"
MUST specify: Language/framework versions, performance requirements, backward compatibility, error handling
Bug Fix:
File: [path] | Error: [exact message]
Expected: [behavior] | Actual: [behavior]
Relevant code: [snippet]
Feature:
Goal: [one sentence]
Criteria: [list] | Location: [file] | Constraints: [limits]
CURRENT.md - Active task state (SHOULD update when switching tasks, MUST reflect current focus)
STATUS.md - Project progress (SHOULD update at session end, MUST include blockers/next steps)
[Start] → [Load CURRENT.md] → [Work + TodoWrite] → [End] → [Update STATUS.md]
Skills SHOULD be: Single-purpose, composable, self-contained
Skills MUST NOT: Conflict with others, duplicate core rules, exceed context budget
| Pattern | Trigger |
|---|---|
| File-based | *.py, Dockerfile, package.json |
| Directory-based | .claude/, tests/, src/ |
| Content-based | Keywords in conversation |
| Command-based | /skill-name invocation |
| Anti-Pattern | Better Approach |
|---|---|
| Context dumping | Provide focused, relevant context |
| Vague requests | Specific, measurable goals |
| Premature optimization | MVP first, optimize with data |
| Ignoring errors | Address errors immediately |
| Anti-Pattern | Better Approach |
|---|---|
| Bash for file ops | Use Read/Edit/Write tools |
| Write over Edit | Prefer Edit for existing files |
| Single-threaded search | Parallel subagent searches |
| Skipping Read | Always Read before Edit |
| Anti-Pattern | Better Approach |
|---|---|
| Coding without tests | TDD or test-after minimum |
| Giant commits | Atomic, logical commits |
| Ignoring TodoWrite | Update status in real-time |
| Anti-Pattern | Better Approach |
|---|---|
| "Make it better" | Define specific improvements |
| Assuming knowledge | Provide relevant background |
| Ignoring constraints | State constraints upfront |