pi-autoresearch
Autonomous experiment loop extension for pi
The Lens
By Erik Loyd, SaaS CEO and former COO/CFO of an AWS Premier Partner.
Updated Aug 2026
Research loops are mostly babysitting: change a parameter, run it, read the output, change it again. Pi-autoresearch takes you out of that loop. You define an experiment, walk away, and it cycles through modify, test, and evaluate until it converges on a result.
This is the same autonomous research pattern as ARIS and autoresearch, but built specifically for pi's ecosystem. The integration is tighter because it is native to the harness rather than a bolt-on. MIT licensed, TypeScript.
Pi itself is an open source agent harness, so there is no platform subscription behind this. The recurring cost is your own model API key, and an autonomous loop that keeps modifying and re-testing will spend more of it than you expect. Cap the iterations before you walk away.
The catch is portability. This is built for pi and only pi. Move to Claude Code or Codex and you leave it behind, which is a real consideration for a tool whose whole value is unattended runtime.
Free vs Self-Hosted vs Paid
fully freeFree
The extension is fully open source under MIT with no paid tier. Pi, the agent harness it extends, is open source as well, so there is no platform subscription layered underneath it.
Self-Hosted Costs
None beyond your own machine. It is a TypeScript extension that runs inside pi.
The Real Cost: Model Tokens
This is the whole bill. An autonomous loop keeps modifying, testing, and evaluating without you in the way, which means it keeps calling the model without you in the way either. Bring your own Anthropic, OpenAI, or compatible key, and set an iteration cap before starting a long run.
When to Pay
Nothing to buy. The decision is how many unattended iterations you are willing to fund.
Free and MIT, and pi itself is open source, so there is no platform fee underneath. Your own model API key is the entire cost, and unattended loops spend it quickly.
What to do by team size
- Solo
- free; cap the iteration count before an unattended run
- Small team
- free; the shared cost is model tokens, not licenses
- Medium team
- free; agree on an iteration budget so loops do not run away
- Large team
- free; pi-only, so weigh the lock-in before standardizing on it
Get tools like this every Wednesday
One featured tool, three on the radar. No fluff.
Similar Tools
ARIS ⚔️ (Auto-Research-In-Sleep): lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, and experiment automation. No framework, no lock-in. Works with Claude Code, Codex, OpenClaw, or any LLM agent.
Fully autonomous & self-evolving research from idea to paper. Chat an Idea. Get a Paper. 🦞
Claude Autoresearch Skill: autonomous goal-directed iteration for Claude Code. Inspired by Karpathy's autoresearch. Modify → Verify → Keep/Discard → Repeat forever.
A low score is not a verdict on quality. Young and niche tools start low by design. How we calculate scores
Trust Signals
License: MIT License
Use freely, including commercial. Just keep the license.
Commercial use: ✓ Yes
About
- Owner
- David Cortés (User)
- Stars
- 8,087
- Forks
- 459
- discussed