Back to directory

Agent-Native

An open-source TypeScript framework for building agents that share a purpose-built UI, so the same action can be called by the agent as a tool or by application code.

Agent-Native

Agent-Native lets a developer define a capability once as an action; the agent uses it as a tool, while the UI calls the identical function from code, so both paths share the same validation, permissions, and implementation. Work the agent does appears in the UI and work done in the UI is visible to the agent, and the agent also receives relevant application state such as the current page or selected record.

Actions are also exposed over HTTP, MCP, and A2A in addition to the CLI. The project ships example open-source agents, such as one for recording and understanding meetings, screens, and voice notes, that can be run directly or adapted as a starting point.

It uses PostgreSQL in production and PGlite for local development, and can be scaffolded with a single CLI command.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.