Back to directory

little-coder

A coding agent tuned for small, locally run language models, built as a curated bundle of extensions and skills on top of the pi agent substrate.

little-coder

little-coder is a coding agent for small, locally run language models, built on top of pi, a minimal agent substrate that supplies the agent loop, a multi-provider API, a terminal UI, session tree, context compaction and an extension model with four built-in tools.

It adds roughly 30 extensions, 30 skill markdown files and a Python benchmark harness on top of plain pi, without forking it: pi stays a normal dependency, and little-coder-specific code lives in its own extensions, skills and benchmarks directories. The launcher starts pi with only its bundled extension set enabled, which keeps the cold-start context small and behavior predictable, since nothing in the working directory changes what loads.

A globally installed pi extension does not load by default for this reason; opting back in is possible through a local extensions folder, an environment variable pointing at extra extension files, or a flag that lets pi discover its own installed extensions.

More in Coding agents

See all 793
  • The City

    github.com

    A native macOS app that shows Claude Code runs as a city, with each project as a building, each team as a floor of robots, and every animation driven by a real event from the claude CLI's stream.

  • rig

    tryrig.ai

    A small, single-binary agent runtime that points at any OpenAI-compatible model, local or hosted, reads and edits a repository's code, runs its tests and remembers what it learned, with a loop implemented in a few hundred lines.

  • Jauvex

    jauvex.reindent.com

    An Electron desktop app that puts Claude Code, Codex and Grok coding-agent sessions side by side and lets you steer them by voice, with a small model handling fast routing decisions.