Task-to-Agent Matcher Skill
Instructions
- Analyze the task to understand requirements
- Identify task type and key requirements
- Review available agents and their capabilities
- Match task requirements to agent capabilities
- Rank agents by suitability
- Recommend best agent(s) for the task
- Identify if new agent is needed
Matching Process
Step 1: Analyze Task
- Read task description, requirements, and acceptance criteria
- Identify task type (Design, Implementation, Infrastructure, Testing, etc.)
- Extract key requirements and capabilities needed
- Note any special requirements or constraints
Step 2: Identify Task Characteristics
- Task Type: Design, Implementation, Infrastructure, Testing, Documentation, etc.
- Domain: Frontend, Backend, Database, DevOps, Security, etc.
- Complexity: Simple, Medium, Complex
- Special Requirements: Accessibility, Performance, Security, etc.
Step 3: Review Available Agents
- List all available agents from
.cursor/agents/directory - For each agent, understand their capabilities
- Use agent-capability-assessor if needed for detailed assessment
Step 4: Match Tasks to Agents
- Compare task requirements to each agent's capabilities
- Score each agent on match quality
- Consider both capability match and workflow fit
Step 5: Make Recommendation
- Rank agents by suitability
- Recommend best agent(s)
- Identify if new agent is needed
- Note any gaps or concerns
Matching Criteria
Task Type Matching
Design Tasks → ui-ux-designer
- UI/UX design work
- Component design
- Layout design
- Design system work
Implementation Tasks → implementation-engineer
- Code implementation
- Feature development
- API endpoints
- Component implementation
Infrastructure Tasks → infrastructure-engineer
- Database setup
- Deployment
- CI/CD
- Environment configuration
Specification Tasks → specification-writer
- Creating specifications
- Requirements analysis
- Technical planning
Task Breakdown → scrum-master
- Breaking down specifications
- Creating task lists
- Dependency management
Code Review → code-reviewer or code-reviewer-feedback
- Reviewing code
- Providing feedback
Testing Tasks → test-runner or test-writer
- Running tests
- Writing tests
Documentation Tasks → documentation
- Creating documentation
- Writing docs
Domain Matching
Frontend → ui-ux-designer, implementation-engineer (for frontend)
Backend → implementation-engineer (for backend), api-implementer
Database → infrastructure-engineer, database-implementer
DevOps → infrastructure-engineer
Security → security-auditor
Accessibility → ui-ux-designer (design), accessibility skills
Matching Output Format
## Task-to-Agent Matching
### Task: [Task ID and Title]
**Type**: [Task Type]
**Domain**: [Domain]
**Complexity**: [Complexity]
### Task Requirements
- [Requirement 1]
- [Requirement 2]
- [Requirement 3]
### Agent Matches
#### 1. [Agent Name] - [Match Quality]
**Match Score**: [Excellent/Good/Fair/Poor]
**Capabilities**:
- [Capability 1]
- [Capability 2]
**Why This Match**:
[Explanation of why this agent matches]
**Gaps/Concerns**:
[Any gaps or concerns]
#### 2. [Agent Name] - [Match Quality]
[Similar format]
### Recommendation
**Best Match**: [Agent Name]
**Alternative**: [Alternative Agent] (if applicable)
**New Agent Needed**: [Yes/No]
**Rationale**: [Why this recommendation]
Examples
Example 1: Clear Match
Input: Match task "TASK-010: Style the submit button"
Output:
## Task-to-Agent Matching
### Task: TASK-010: Style the submit button
**Type**: Implementation (Design needed first)
**Domain**: Frontend
**Complexity**: Simple
### Task Requirements
- Create button styling
- Implement design specifications
- Ensure accessibility
- Add responsive design
### Agent Matches
#### 1. ui-ux-designer - Excellent Match
**Match Score**: Excellent
**Capabilities**:
- Creates design specifications
- Designs UI components
- Considers accessibility
- Plans responsive design
**Why This Match**:
This task needs design specifications first. ui-ux-designer specializes in creating design specs for components like buttons, including accessibility and responsive considerations.
**Gaps/Concerns**: None
#### 2. implementation-engineer - Good Match (after design)
**Match Score**: Good
**Capabilities**:
- Implements design specs
- Handles frontend implementation
- Ensures code quality
**Why This Match**:
After design is complete, implementation-engineer would implement the button styling according to design specs.
**Gaps/Concerns**: Requires design specs first
### Recommendation
**Best Match**: ui-ux-designer (for design), then implementation-engineer (for implementation)
**New Agent Needed**: No
**Rationale**: This task requires both design and implementation. ui-ux-designer should create design specs first, then implementation-engineer should implement them.
Example 2: Multiple Options
Input: Match task "TASK-015: Set up PostgreSQL database with user and task tables"
Output:
## Task-to-Agent Matching
### Task: TASK-015: Set up PostgreSQL database with user and task tables
**Type**: Infrastructure
**Domain**: Database
**Complexity**: Medium
### Task Requirements
- Set up PostgreSQL database
- Create database schema
- Create user and task tables
- Set up migrations
- Configure database connection
### Agent Matches
#### 1. infrastructure-engineer - Excellent Match
**Match Score**: Excellent
**Capabilities**:
- Handles infrastructure tasks
- Sets up databases
- Creates database schemas
- Handles deployment and configuration
**Why This Match**:
infrastructure-engineer specializes in infrastructure tasks including database setup, schema creation, and configuration.
**Gaps/Concerns**: None
#### 2. database-implementer (skill) - Good Match
**Match Score**: Good
**Capabilities**:
- Creates database schemas
- Handles migrations
- Creates data access layers
**Why This Match**:
The database-implementer skill provides database-specific capabilities, but would need to be used by an agent like infrastructure-engineer.
**Gaps/Concerns**: This is a skill, not an agent. Needs to be used by an agent.
### Recommendation
**Best Match**: infrastructure-engineer
**New Agent Needed**: No
**Rationale**: infrastructure-engineer is the best match for database infrastructure tasks. It can use the database-implementer skill for detailed database work.
Example 3: New Agent Needed
Input: Match task "TASK-020: Create machine learning model for image classification"
Output:
## Task-to-Agent Matching
### Task: TASK-020: Create machine learning model for image classification
**Type**: Implementation
**Domain**: Machine Learning / AI
**Complexity**: High
### Task Requirements
- Design ML model architecture
- Train image classification model
- Evaluate model performance
- Deploy model
### Agent Matches
#### 1. implementation-engineer - Poor Match
**Match Score**: Poor
**Capabilities**:
- General code implementation
- API and backend work
- Standard software development
**Why This Match**:
Implementation-engineer handles general implementation but doesn't specialize in machine learning.
**Gaps/Concerns**:
- Lacks ML-specific knowledge
- No ML model training experience
- No ML deployment expertise
#### 2. No other existing agents match
### Recommendation
**Best Match**: None (new agent needed)
**New Agent Needed**: Yes
**Rationale**: This task requires specialized machine learning capabilities that no existing agent has. A new ML-engineer or data-scientist agent should be created to handle this task effectively.
Multi-Task Matching
When matching multiple tasks:
- Match each task individually
- Consider agent workload
- Optimize assignments
- Identify patterns
## Multi-Task Agent Matching
### Tasks to Match
- TASK-001: [description]
- TASK-002: [description]
- TASK-003: [description]
### Task Assignments
- TASK-001 → [agent-name]
- TASK-002 → [agent-name]
- TASK-003 → [agent-name]
### Agent Workload
- [agent-name]: [number] tasks
- [agent-name]: [number] tasks
### Recommendations
[Overall recommendations and patterns]
Best Practices
- Be Thorough: Analyze task requirements completely
- Consider Workflow: Match workflow fit, not just capabilities
- Think About Dependencies: Consider task dependencies when matching
- Optimize Assignments: Balance workload across agents
- Identify Gaps: Clearly identify when new agents are needed
- Document Rationale: Explain why each match was chosen
Scan to join WeChat group