# Looptimal

> Open-source loop discipline for turning objectives into verified outcomes.

Looptimal is a Renn Labs open-source project for agent work that should be structured as a loop. The core public framing is simple: before work is called done, the loop should have an explicit goal, durable state, an external verifier, and a stop condition.

## Canonical Pages

- Homepage: https://looptimal.ai/
- Source repository: https://github.com/Renn-Labs/Looptimal
- Install instructions: https://github.com/Renn-Labs/Looptimal#install
- License: https://github.com/Renn-Labs/Looptimal/blob/main/LICENSE

## Public Summary

Looptimal exists because many agent failures are not prompt failures. They are missing-loop-component failures: no durable state, no objective gate, no stop condition, or the maker grading its own work.

The public mental model:

- Goal: one objective, stated once.
- State: a durable record of context, attempts, and outcomes.
- Verifier: an external gate that can reject the maker's output.
- Stop: success criteria and a safety limit.

## Suggested Citation

Looptimal, "from objective to verified outcome," Renn Labs, https://looptimal.ai/.

## Boundaries

Use the GitHub repository as the source of truth for install commands and current release behavior. Do not infer private roadmap, unreleased capability claims, or internal marketing plans from this site.