Your Automated Agent Builder, Right on Your Desktop / Server
Create Self-Evolving Agents in One Click
macOS · Windows · Linux
English | 简体中文
With LangChain, you build agents by hand — at 1× speed.
With PenguinHarness, agents build agents — at 100×.
Three reasons, in deliberate order — from task quality, to how agents get built, to how they keep improving.
A deliberately minimal toolset over clean low-level interfaces: fewer tool calls, fewer tokens — deeply tuned for open models like DeepSeek. Each harness on the model it is normally paired with, same tasks, head-to-head:
Best accuracy on data analysis — at 1/70 of Claude Code's cost.
Type one sentence, and an agent builds the complete agent application for you — scaffold, code, and run instructions, end to end:
Collect the docs from https://github.com/ericbuess/claude-code-docs and build a RAG app that answers Claude Code questions as a configuration expert, citing its sources.
And this is the finished product — a docs expert with retrieval, cited sources that link to the original files, and example questions built in:
rag_en.mp4
And generating this entire RAG app burned just $0.02 (¥0.2) of tokens — on DeepSeek V4 Pro.
With PenguinHarness Skills, an agent evaluates and optimizes itself: run the benchmark, find the lost points, ship version N+1 — with a snapshot before every round, and every request observable in the Trace view.
evo_en.mp4
Four Skill groups ship in the box (docs); agents can also write and optimize their own:
| Group | Skills |
|---|---|
| Office Productivity | data-analysis, firecrawl |
| Software Development | web-design, software-engineering |
| AI App Development | penguin-sdk, penguin-cli, agenthub-models, vllm, ollama, llamafactory |
| Agent Tuning | agent-creation, benchmark-design, agent-evaluation, agent-optimization |
| Model | Providers |
|---|---|
| DeepSeek V4 | DeepSeek, OpenRouter, Fireworks AI, SiliconFlow, Qwen Token Plan, Qwen Pay-As-You-Go |
| Kimi K3 | Moonshot AI, OpenRouter, Qwen Pay-As-You-Go |
| GLM 5.2 | Z.AI, OpenRouter, Fireworks AI, SiliconFlow, Qwen Token Plan, Qwen Pay-As-You-Go |
| Hunyuan 3 | OpenRouter |
| Qwen 3.8 Max | Qwen Token Plan, Qwen Pay-As-You-Go, OpenRouter |
| GPT 5.6 | OpenRouter |
| Gemini 3.6 Flash | Google Gemini, OpenRouter |
| Claude 5 | Anthropic, OpenRouter |
| Inkling | OpenRouter, Fireworks AI |
Each family's latest generation only — the app's Models page lists every built-in preset, and any OpenAI-protocol endpoint works too: pick a preset, or point a custom endpoint at any of the 1000+ online and local models.
| Requirement | Supported |
|---|---|
| OS | Linux, macOS, Windows 10+ |
| Architecture | x64, arm64 |
| Runtime | bundled by the one-line installer (npm installs need Node >= 24) |
| Model | an API key for at least one model |
Two ways in — both work on the same ~/.penguin/data root, so a desktop install and a CLI install can be mixed freely:
- 🖥️ Desktop app — a double-click install: it embeds the server and opens already signed in, no terminal involved.
- ⌨️ CLI — a one-line installer (or npm / offline package) puts the
penguincommand on the machine;penguin webthen serves the full Web experience in your browser athttp://127.0.0.1:7364(multi-session chat, agent / skill / model management, usage stats, Trace observability, evaluation center). The online installers bundle their own Node runtime — unpack and run; upgrades and reinstalls never touch your data.
Note
On a CLI install, the first Web login is admin, with the initial password printed on the server's first start (of the form penguin-1234) — change it right after. Models are configured on the in-app Models page.
The full Web experience as a standalone application: it embeds the server and opens already signed in — no terminal, no login page, no initial password to copy. It works on the same ~/.penguin/data root as a CLI install, so the two can be used interchangeably (a data root only ever runs one server; if a CLI-started instance is already up, the app attaches to it).
⬇️ Get it from the download page — the page serves the OSS-accelerated mirror when it is reachable, and every installer is also attached to each GitHub Release.
| Platform | Installers |
|---|---|
| macOS 11+ | dmg (Apple Silicon / Intel) |
| Windows 10+ | installer (.exe, x64) |
| Linux (x64) | AppImage / deb |
Current builds are unsigned, so the system may block the very first launch. Expand your platform for the one-time fix:
🍎 macOS says “PenguinHarness” is damaged and can’t be opened
macOS quarantines files downloaded from the internet, and the missing signature makes that flag surface as a false “damaged” alert. Deleting the flag clears it:
-
Open the downloaded dmg and drag
PenguinHarness.appinto the Applications folder. -
Open Terminal (Launchpad → Other → Terminal).
-
Paste this command into Terminal and press Enter, then type your login password (nothing shows while you type; press Enter when done):
sudo xattr -rd com.apple.quarantine /Applications/PenguinHarness.app
-
Once it finishes, double-click the app — it now opens normally.
🪟 Windows SmartScreen says “Windows protected your PC”
The installer is not signed yet, so SmartScreen holds the first run: click More info, then Run anyway to continue installing — first run only.
🐧 Linux: double-clicking the AppImage does nothing
Browsers download AppImages without the execute permission. Grant it once and the app starts normally from then on (the deb package installs through the package manager and is not affected):
chmod +x penguin-desktop-linux-x86_64.AppImagecurl -fsSL https://penguin.ooo/install.sh | sh
penguin web # start the service and open http://127.0.0.1:7364irm https://penguin.ooo/install.ps1 | iex
penguin web # start the service and open http://127.0.0.1:7364npm install -g @prismshadow/penguin-cli
penguin web # start the service and open http://127.0.0.1:7364📴 Offline install (air-gapped machines)
Every GitHub Release attaches exactly one package per target — Linux and macOS in x64 / arm64, Windows in x64, plus a runtime-less universal package — and the same file serves online and offline installation. Each package seals the program payload, its SHA256 checksum and the platform's installer: download the one file on a networked machine, copy it to the target, extract once and run the bundled installer — no network, no separate checksum file to carry (the sealed SHA256 is always verified).
Linux (on arm64, use penguin-linux-arm64.tar.gz):
mkdir penguin-install
tar -xzf penguin-linux-x64.tar.gz -C penguin-install
./penguin-install/install.shmacOS (Apple silicon shown; on Intel, use penguin-darwin-x64.tar.gz):
mkdir penguin-install
tar -xzf penguin-darwin-arm64.tar.gz -C penguin-install
./penguin-install/install.shWindows (unzip, then double-click install.cmd — or run it in PowerShell):
Expand-Archive penguin-win32-x64.zip -DestinationPath penguin-install
cd penguin-install
.\install.cmdThe same engine, scriptable — made to be driven by agents (and agents building agents):
penguin config model add --provider deepseek --model-id deepseek-v4-flash --api-key sk-... --set-default
penguin run -m "Create hello.txt containing Hello, Penguin" # one-shot task
penguin chat # interactive REPL (/compact, /exit, Ctrl-C to interrupt)
penguin server # headless service (same API the Web App uses)import { createAgent, isCompleteModelMessage, userText } from "@prismshadow/penguin-core";
const agent = await createAgent({ agentId: "default_agent" });
const session = await agent.createSession({ workspaceDir: process.cwd() });
for await (const output of session.run([userText("Create hello.txt containing hi")], {
approve: async () => "allow", // per-tool-call approval
})) {
if (isCompleteModelMessage(output) && output.payload.type === "text") {
console.log(output.payload.text);
}
}- Public release of the benchmark suite
- Desktop app
- Windows support
- Agent company and templates
- Company-level self evolving
- OpenShell integration (permission-governed shell)
- More to come…
pnpm install && pnpm build # build first: core's exports point at dist/
pnpm dev # backend + web app together (prefixed logs, deps built once)See CONTRIBUTING.md for the full workspace guide: dev commands, quality gates, repo layout, and the changelog rule.
Thanks to everyone who has contributed to PenguinHarness!
If you use PenguinHarness in your research, please cite:
@software{penguinharness2026,
author = {{PrismShadow Team}},
title = {PenguinHarness: Efficient Self-Improving Harness for Everyone},
year = {2026},
url = {https://github.com/Prism-Shadow/penguin-harness},
license = {Apache-2.0}
}Apache-2.0 © 2026 Prism Shadow
Built with ❤️ by Yaowei Zheng (author of LlamaFactory), the PrismShadow AI Team, and Fable 5.