what you get
registry + runtime + tools
Discovery, validation, quality scoring, sandboxed execution, MCP endpoints, collections, agents, and an SDK. All open source.
tool registry
100+ AI tools from npm, auto-discovered within minutes of publication. Quality scored and health monitored.
omega agent
An AI agent that discovers and runs tools on the fly based on what you ask. Zero config.
collections
Group tools into workflow bundles. Add test scenarios to validate behavior and auto-generate usage docs.
custom agents
Build agents with any LLM, custom prompts, and curated tool sets. Share publicly or keep private.
mcp protocol
One URL gives Claude Desktop, Cursor, Windsurf, or any MCP client instant access to your tools.
secure execution
Tools run in isolated sandboxes with rate limiting and timeouts. Credentials encrypted at rest.
test scenarios
Auto-generated test scenarios validate tool behavior. Track pass rates, latency, and quality scores.
living skills
Documentation that evolves from real usage. Skills surface from question patterns and proven tool combinations.
developer sdk
One npm keyword to publish. Zod schemas auto-extracted, Vercel AI SDK format, TypeScript-first.