# Prism-Shadow/penguin-harness

> A framework for agent self-evolution (RSI): the model stays fixed while the harness evolves — prompts, Skills, tools and hooks. Runs as a desktop app, Web App, CLI and SDK with algorithm plugins (OPRO, APE, ACE, AWM) and built-in benchmarks.

- Magnitude: 1.6 out of 10 — Cooling
- Stars: 2,988 total · +3 stars measured 2026-10-27, 09:12–11:01 UTC, ≈ 10 by evening
- Star trust: star growth looks organic
- Category: AI agents · Language: TypeScript · License: Apache-2.0 · Created: 2026-07-19 · Last push: 2026-10-27
- GitHub: https://github.com/Prism-Shadow/penguin-harness · Homepage: https://penguin.ooo/ · Page: https://gitnova.dev/en/r/Prism-Shadow/penguin-harness

## Useful for

- Run an RSI loop so one agent improves another against a chosen benchmark
- Build an agent app from a single text description of the task
- Compare OPRO, APE, ACE and AWM algorithms on the same benchmark via the Evaluation Center

## Why it’s here

- Star-counter measurements on 2026-10-27 (UTC), 09:12–11:01: 2985 → 2988 stars (+3). This is the change over that interval.
- Estimated end-of-day forecast: about +10 stars, using observed gains and the previous day.
- Interest is fading: the two-day pace is 36% of the previous week and a half.

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 316
- Issues and pull requests: 1,371
- Watchers: 17
- Average over the last week: 14 per day
- Usual pace: 33 per day
- Stars in the last hour (measured): 1
- Latest release: v0.2.18 (2026-10-27)

## Stars per day, last 30 days (oldest → newest, today is partial)

2026-09-28 … 2026-10-27: 15, 8, 5, 5, 3, 3, 4, 5, 2, 5, 7, 8, 8, 11, 12, 15, 28, 19, 218, 36, 36, 26, 26, 20, 11, 12, 9, 20, 19, 3

## Similar by description

1. **Human-Agent-Society/reef** — 1.8 · Steady · AI agents · Python · +6 stars measured 2026-10-27, 00:40–11:08 UTC, ≈ 14 by evening
   Continual learning infrastructure for self-improving agents: connects agent inference, feedback collection, learning, and versioned delivery of updates. Lets you train model weights or improve an agent's harness (prompts, rules, skills).
   Full card: https://gitnova.dev/en/r/Human-Agent-Society/reef.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-27 11:09 UTC, updated every 30 minutes.