Features
Autonomous mission dispatch with milestones and parallel workers
Pulse supervision for scheduled checks and automated PR merging
Domain-specific agents for code, ops, marketing, legal, etc.
Token-efficient context compression with .agents folder
Worktree-based parallel development with isolated branches
Cross-model review for high-risk operations
Built-in secret handling and supply-chain security scans
Budget tracking and token cost optimization
Imported skills and session learnings
Integration with 30+ third-party services
Slash commands and MCP tooling
Observability signals for stuck workers and CI failures
Support for OpenAI GPT-5.5, GPT-5.4 mini, Anthropic, Google, local models
OpenCode integration with model-account pools
TOON-ready context for 20-60% token reduction
@ai_fn decorator for LLM-powered Python functions
@ai_classifier decorator for text classification
Structured data extraction via Pydantic models
Agent loops with tool calling and function calling
Built-in streaming (SSE) support
Async-first API for concurrent applications
Rate limiting and automatic retries
LLM response caching for performance
Concurrency control for workload management
CLI tool for monitoring and debugging
SQLite state store for persistence
OpenAI and Anthropic model support
Embeddings generation for semantic search
Local execution with no cloud dependency
Self-hosted deployable as a library
