Skills Overview¶
Quaestor provides Skills - auto-activating workflows that handle specific development tasks. Skills trigger automatically based on context and orchestrate sub-agents internally.
What are Skills?¶
Skills are specialized workflows that: - Activate automatically based on keywords and context - Orchestrate multiple sub-agents internally - Provide focused capabilities for specific tasks - Work through natural language - no manual invocation needed
Available Skills¶
Quaestor includes 7 Skills:
1. managing-specifications¶
Activates when: Creating specs, checking status, managing lifecycle
What it does: - Creates specifications from requirements - Manages spec lifecycle (draft → active → completed) - Tracks progress from checkboxes - Enforces 3-active-spec limit
Example triggers: - "Create a spec for user authentication" - "Show my active specs" - "Activate spec-feature-001" - "Complete spec-auth-001"
2. implementing-features¶
Activates when: Using /implement or implementing features
What it does: - Loads specifications and requirements - Coordinates research, architecture, implementation, testing, review - Tracks progress via TODOs - Runs tests and validates criteria
Example triggers:
- /implement spec-auth-001
- /implement "Add rate limiting"
3. reviewing-and-shipping¶
Activates when: Creating PRs, shipping work
What it does: - Validates code quality - Checks test coverage - Auto-fixes issues - Generates commit messages - Creates GitHub PRs with context
Example triggers: - "Create a pull request for spec-feature-001" - "Ship this code" - "Review and create PR"
4. debugging-issues¶
Activates when: Debugging issues, investigating errors
What it does: - Systematic root cause analysis - Error reproduction steps - Hypothesis testing - Fix implementation with tests
Example triggers: - "Debug the login failure" - "Investigate why the API is slow" - "Fix the memory leak"
5. security-auditing¶
Activates when: Security analysis, vulnerability detection
What it does: - Scans for vulnerabilities - OWASP Top 10 checking - Authentication/authorization review - Input validation analysis
Example triggers: - "Review this code for security issues" - "Audit authentication security" - "Check for SQL injection"
6. optimizing-performance¶
Activates when: Performance improvements, profiling
What it does: - Profiles code execution - Identifies bottlenecks - Implements caching strategies - Database query optimization
Example triggers: - "Optimize database queries" - "Profile the API endpoint" - "Improve response time"
7. initializing-project¶
Activates when: Setting up Quaestor in projects
What it does:
- Auto-detects project framework
- Creates .quaestor/specs/ structure
- Configures appropriate settings
- Adaptive setup based on project type
Example triggers: - "Initialize Quaestor" - "Setup this project"
How Skills Work¶
Automatic Activation¶
Skills activate based on context and keywords:
User says: "What's the status of spec-auth-001?"
Behind the scenes:
1. Claude detects keywords: "status", "spec"
2. managing-specifications skill activates
3. Reads .quaestor/specs/ folders
4. Calculates progress from checkboxes
5. Reports status to user
You see: "spec-auth-001: 80% complete (4/5 criteria)"
Sub-Agent Orchestration¶
Skills coordinate multiple sub-agents internally:
Example: implementing-features
/implement spec-auth-001
Skill orchestrates:
1. researcher → analyze existing patterns
2. architect → design approach
3. implementer → write code
4. qa → create tests
5. reviewer → validate quality
You don't call agents directly - Skills handle all orchestration.
Tool Restrictions¶
Skills have controlled tool access for safety:
managing-specifications: Write, Read, Edit, Bash, Glob, Grepimplementing-features: All tools (full implementation capability)reviewing-and-shipping: Read, Write, Edit, Bash, Grep, Globdebugging-issues: Read, Edit, MultiEdit, Bash, Grep, Globsecurity-auditing: Read, Grep, Glob, Bash, Task, WebSearchoptimizing-performance: Read, Edit, Bash, Grep, Glob, Taskinitializing-project: Read, Write, Edit, Glob, Grep
Benefits¶
For Users: - Zero learning curve - just describe what you want - No manual coordination - skills activate automatically - Natural language interface - Always available via plugin
For Development: - Maintainable - skills are markdown files - Composable - skills can work together - Safe - tool restrictions prevent accidents - Discoverable - users can read SKILL.md files
Skills vs Sub-Agents¶
| Feature | Skills | Sub-Agents |
|---|---|---|
| Activation | Automatic (keyword-triggered) | Internal (skill-invoked) |
| Invocation | Natural language | Task tool by skills |
| Purpose | Workflow orchestration | Specialized technical work |
| Tool access | Restricted by allowed-tools | Determined by skill |
| User interaction | Direct | Indirect (via skills) |
Key difference: You interact with Skills, Skills orchestrate sub-agents.
Complete Workflow Example¶
1. User: "Create a spec for user authentication"
→ managing-specifications: Creates spec in draft/
2. User: "Activate spec-auth-001"
→ managing-specifications: Moves to active/
3. User: "/implement spec-auth-001"
→ implementing-features: Orchestrates research/design/code/test/review
4. User: "What's the progress?"
→ managing-specifications: "80% complete (4/5 criteria)"
5. User: "Complete spec-auth-001"
→ managing-specifications: Moves to completed/
6. User: "Create a pull request"
→ reviewing-and-shipping: Validates quality and creates GitHub PR
All natural language - no commands to memorize!
Learn More¶
- /plan Command - Activates managing-specifications
- /implement Command - Activates implementing-features
- Sub-Agents - Internal agents orchestrated by skills
- Specification-Driven Development - Core workflow
Skills make Quaestor effortless - describe what you want, and it happens automatically.