🐧 Unified and Stable RSI Platform
PenguinHarness is a local-first platform in which multiple AI agents create, evaluate, optimize, and deploy agent applications. It is for people building AI software who want agents to generate applications and improve their own behavior through skills.
This repository also configures its own agents. See what penguin-harness tells them →
Skill Claude CodeCodex
Run experiments and papers in a PenguinHarness research organization — fix the harness and the metric first, run an autoresearch-style loop (one editable surface, the same time budget per experiment, a results log, keep only improvements) inside a resource envelope the board approved in the all-hands channel, and put…
Skill Claude CodeCodex
Drive the PenguinHarness agent browser — the desktop app's built-in browser or the user's own Chrome — from the shell with penguin browser: open pages, read them as simplified HTML or text, act with JavaScript and trusted clicks and typing, and pull structured data out of them (orders, search results, tables), signed…
Skill Claude CodeCodex
Make a reply easier to read and act on with rich blocks inside ordinary Markdown — a choice the user picks from, a form that collects several answers, a procedure as steps with warnings in place, a callout, a Mermaid diagram, a weather card, a live clock or countdown, and a dashboard of gauges for a system snapshot, a…
Skill Claude Code
Turn an external plugin — a Codex or Claude Code plugin, a skills repository, a folder in a GitHub repository — into a PenguinHarness plugin package, its skills and its MCP servers, and install it on the server's plugin market with penguin plugin install ; the package format, the mapping rules and the review duties.
Skill Claude CodeCodex
Run ACE (Agentic Context Engineering, Zhang et al. 2025) on an agent the sequential offline way - a Reflector reads each training Trace, a Curator adds the lessons to a playbook of rules the agent reads before every task, and the final playbook is measured on a published Benchmark.
Skill Claude CodeCodex
Use for workbenches — the tabs and pages beside the chat, with server code behind them, that an Agent keeps in its own workbenches/ folder inside PenguinHarness — building one, changing it, showing it as the whole app, restoring an earlier version, and removing a tab or a whole workbench when the user wants the custom…
Skill Claude CodeCodex
Reproduce APE (Large Language Models are Human-Level Prompt Engineers, Zhou et al. 2022) on an agent and a frozen Benchmark, or on the house-style brief demo task shipped with this Skill: induce instructions from input/output examples, measure each one, keep the best.
Skill Claude CodeCodex
Run AWM (Agent Workflow Memory, Wang et al. 2024) on an agent - induce reusable workflows from its successful trajectories, online over a training stream or offline from supplied experiences, then freeze the workflow memory and measure it on a published Benchmark.
Skill Claude CodeCodex
Reproduce OPRO (Large Language Models as Optimizers, Yang et al. 2023) on an agent and a frozen Benchmark, or on the precise-summary demo task shipped with this Skill: propose instructions from the scored history, measure each one, keep the best.
Skill Claude CodeCodex
Answer PenguinHarness product questions and troubleshoot setup, models, agents, plugins and workbenches using the official documentation and source code. Use for every question handled by the PenguinHarness Support agent.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: