33 expert agents for autonomous task execution
// Spawn an Engineer agent for implementation
await Task({
subagent_type: "Engineer",
description: "Build user authentication system",
prompt: "Implement JWT-based auth with refresh tokens"
});
// Research agent for web investigation
await Task({
subagent_type: "GeminiResearcher",
description: "Research latest security practices",
prompt: "Find best practices for OAuth 2.0 implementation"
});
Versatile agents for general-purpose work
Default agent for general-purpose tasks. No special tools or constraints. Handles a wide variety of work.
Junior-level agent for simple, well-defined tasks. Cost-effective for straightforward work.
Exploratory agent for investigation and discovery. Good at finding patterns and insights.
Product-focused agent for requirements, planning, and user stories. Translates needs into specifications.
Specialized agents for software engineering
Senior engineering leader for strategic implementation. Emphasizes TDD, comprehensive planning, constitutional compliance.
UX/UI design specialist. Creates user interfaces, wireframes, and design systems following best practices.
System architecture and design patterns. Strategic technical decisions for scalable systems.
Multi-agent code review system. Comprehensive analysis from multiple perspectives for code quality.
Specialized agents for information gathering and analysis
Claude-powered research agent. Deep analysis, synthesis, and comprehensive research workflows.
Gemini-powered web research agent. Fast factual lookup and latest documentation retrieval.
GPT-4 powered research via Codex CLI. Code-focused research and technical analysis.
Perplexity-powered research agent. Real-time web search with citations and source tracking.
Specialized agents for security and infrastructure
Penetration testing specialist. Security assessments, vulnerability discovery, threat modeling.
Incident response coordination. Manages security incidents, coordinates remediation, tracks findings.
Database operations specialist. Schema design, migrations, query optimization, backup management.
Infrastructure operations. Server management, deployment automation, monitoring, scaling.
Agents focused on testing and quality verification
Quality assurance testing. Test plan creation, execution, bug reporting, regression testing.
Reality check agent. Validates assumptions, identifies flaws, challenges design decisions.
UI/UX review specialist. Heuristic evaluation, accessibility checks, design consistency validation.
Unique agents for specific workflows
MCP server builder. Creates Model Context Protocol servers, tool definitions, integration code.
Multi-agent coordination. Orchestrates teams of agents, manages dependencies, coordinates parallel work.
PAI Algorithm executor. Runs the full OBSERVE → PLAN → EXECUTE cycle for complex multi-step work.