Skillget

piloop

agent
Security Audit
Warn
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 5 GitHub stars
Code Pass
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

Autoresearch for any repo: coding agents try ideas, piloop scores every commit and keeps what wins

README.md

piloop

CI
License: MIT
Python 3.14

Autoresearch for any git repo: coding agents try ideas, piloop scores what they submit, and the best commit is the result.

Documentation: piloop.autoresearch.inc

piloop runs coding agents on a goal that a program can measure. Each agent tries one idea in its own git worktree
and submits the commits that it would ship. piloop scores each with your evaluate command into a pool, and the best
commit in the pool is the campaign's result.

piloop watch on the enwik8 example: 10 Claude Code runs took the compressed size from 3,688,770 to 2,361,893 bytes

The enwik8 example after 10 runs: Claude Code agents took the first 10 MB of Wikipedia from 3,688,770 bytes (zlib) to 2,361,893 bytes (−36%) in 1.6 hours.

piloop keeps its state on your machine: a SQLite database in ~/.piloop and git refs in your repo. It has no hosted
service. The agents send your code and prompts to their model providers, as they do when you use them yourself.

Start

  1. Install piloop and its skills for your coding agents:

    curl -fsSL https://raw.githubusercontent.com/hiloopai/piloop/main/install.sh | sh

  2. Open your coding agent in your repo: Claude Code, Codex, pi, Gemini CLI or Cursor.

  3. Tell it what to improve. Say "piloop", so that the agent uses the piloop skills:

    Use piloop to make our tokenizer faster without changing its output.

The agent sets up the campaign, runs it, and asks you only what needs you. To try piloop on an example first, ask for
"the enwik8 example campaign from piloop", or run it by hand.

Requirements

  • Linux (x86-64 or ARM, glibc 2.28 or later) or macOS on Apple silicon. On an Intel Mac, installing builds
    covalence from source, which needs a recent Rust.
  • git 2.40 or later, with a commit identity.
  • uv. The installer installs it.
  • At least one of Claude Code, Codex or
    pi, logged in.

Documentation

  • Quickstart: your first campaign, through your agent.
  • Concepts: campaigns, ideas, runs and the pool.
  • Plugins: add a best rule, a runtime, tables or commands.
  • Reference: the CLI, the configuration and the database schema.

Contributing

See CONTRIBUTING.md. Agents that operate piloop
start at AGENTS.md.

License

MIT. See LICENSE.

Reviews (0)

No results found