# Prism-Shadow/penguin-harness

🐧 Harness for RSI. Let AI Build AI. Multi-Agent Auto-Dev Platform. Everything is Transparent.

Repository: https://github.com/Prism-Shadow/penguin-harness
Canonical: https://ross.abutalabs.com/products/penguin-harness
Homepage: https://penguin.ooo/
Language: TypeScript
License: Apache-2.0
License Family: permissive
Topics: agent, agentic-ai, ai, claude-code, llm, harness, build-tool, desktop, deepseek, rsi, self-evolving, deepseek-harness
Last push: 2026-08-26T20:01:27+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 3
- inputs: {"age_days": 45, "days_push": 7, "days_rel": 7, "gap_med": 2, "n_releases_24m": 14}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1751, forks 181 (observed 2026-08-28T04:05:30.060712+00:00)

## What it is
PenguinHarness is an open-source, local-first multi-agent application development platform that automates building, optimizing, and deploying AI applications. It positions itself as a self-improving harness where AI agents autonomously build and evolve other AI systems, supporting 1000+ models.

## Use cases
- automate building AI applications with multiple agents
- let AI agents build and improve other AI agents
- orchestrate multi-agent auto-development workflows locally
- self-evolving agent harness for recursive self-improvement
- build and deploy LLM-powered apps without manual coding
- run a transparent local-first agent development platform
- optimize and deploy AI apps automatically from a desktop app

## When to choose
- you want a desktop, local-first platform for multi-agent AI app development
- you want agents to autonomously generate, optimize, and deploy AI applications
- you need model flexibility across 1000+ LLM providers
- you want an Apache-2.0 licensed self-evolving agent harness

## When to avoid
- you need a simple single-agent chatbot or prompt library
- you require a fully managed cloud service rather than a local desktop app
- you are uncomfortable with experimental recursive self-improvement (RSI) workflows
- you need a lightweight CLI-only tool rather than a full platform

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, llm-inference, build-tool, developer-tools, workflow-automation
- domain: large-language-models, developer-tools, artificial-intelligence
- platform: cross-platform
- tags: multi-agent, self-evolving, rsi, harness, local-first, claude-code, auto-dev, llm-orchestration, ai-agents, automation, desktop, nodejs, typescript

## Member repositories
- Prism-Shadow/penguin-harness (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:30.060712+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:28:28.030542+00:00, confidence not recorded.
  - readme: https://github.com/Prism-Shadow/penguin-harness (fetched 2026-08-28T04:05:30.060712+00:00, sha 4305c4ef05c9)
  - homepage: https://penguin.ooo/ (fetched 2026-08-29T11:06:43.020110+00:00, sha 3af7031ff3d9)
- Data as of 2026-08-30T08:39:29.467469+00:00.