Changelog, dev startup, README, AgentHub 0.4.0, model catalog, and landing site (#7)

Branch-length batch covering tooling, the model layer, the Web App and the public surfaces. Highlights: - Changelog: a per-release `changelog/<version>/` tree, grouped by the surface each change touches, with a root CHANGELOG.md holding one line per release. - Dev startup: `scripts/dev-prebuild.mjs` serializes the skills+core prebuild behind a lock and keeps `pnpm install` current; `pnpm dev` runs server+web together. - AgentHub 0.3.3 -> 0.4.0: OmniMessage complete payloads carry one opaque `fidelity` object in place of item-level `signature`/`phase`, threaded verbatim through Trace, replay and resume; malformed classification adapted to the new error types. - Model layer: a model is always referenced by an explicit `(provider, model_id)` pair. The provider is never inferred, guessed or defaulted -- both the catalog inference and the unique-match config resolution are gone, and CLI, SDK, server routes and run_subagent all require the complete pair. Catalog gains the Qwen Token Plan, Qwen Pay-As-You-Go and Fireworks AI gateways, plus an expanded OpenRouter group. - Web App: catalog preset sync and per-group speed test on the Models page, positional slash commands, a markdown renderer, skill-library update reminders, and a vertically centred draft page whose upward menus size themselves to the room available. - Public surfaces: restructured READMEs, the penguin.ooo landing site and blog, refreshed benchmark results for both suites, and the demo videos playing on the landing page. Includes the fixes from a full review of the branch: 23 confirmed findings, among them a provider-inference bug that could send one vendor's API key to another vendor's endpoint, and an Escape handler that destroyed the composer's contents unrecoverably. Verified on the branch head: pnpm test (1127 passing, 7 packages), pnpm typecheck and pnpm format:check clean, Playwright e2e 14/14.
2026-07-21 17:43:31 +08:00
1 parent abf0a2f248
commit d4faee3a1e
@@ -0,0 +1,84 @@
+---
+title: "Partnering with the AMD Developer Program: $50 in free Fireworks credits, wired into PenguinHarness"
+date: 2026-07-20
+category: news
+excerpt: We are delighted to partner with the AMD AI Developer Program to bring you free Fireworks coupon codes — apply for the $50 credits with this guide, then set them up in PenguinHarness in three steps.
+---
+
+We are delighted to partner with the **AMD AI Developer Program** to bring you free Fireworks redemption codes: join the program, pass the review, and you receive a Coupon Code redeemable for **$50 in Fireworks AI Credits**. PenguinHarness ships a built-in Fireworks AI gateway group — OpenAI protocol, preset base URL, five preset models — so the credits are usable the moment they land.
+
+> Page content, credit amounts, and expiration dates are subject to change. Please refer to the application page and approval email for the latest information.
+
+## Step 1: Join the AMD AI Developer Program
+
+Visit the official page: <https://developer.amd.com/ai-developer-program/>. In the **Join the AMD AI Developer Program** section:
+
+- No AMD ADP account yet: click **Create Account** and enter your personal information;
+- Already have one: click **Log In** and follow the on-screen instructions.
+
+![The AMD AI Developer Program join page](https://github.com/user-attachments/assets/47a3055b-9a95-40a1-80c6-e3bac7a9ac49)
+
+## Step 2: Request Cloud Credits under Member Perks
+
+After registering and signing in:
+
+1. Click **Member Perks** in the top navigation bar;
+2. Find **Cloud Credit Options**;
+3. Click **Request Cloud Credits** at the bottom of the page.
+
+![Cloud Credit Options under Member Perks](https://github.com/user-attachments/assets/773f8cf1-d72f-4aa3-83b8-f31fc7c9ed9e)
+
+## Step 3: Complete the application form
+
+Fill in the requested personal information:
+
+- Under **Product Needed**, select **Fireworks AI**;
+- In the **Profile** section, provide at least one public profile for account verification: LinkedIn, GitHub, a portfolio, a company or school profile, WeChat, WeCom, and similar all work.
+
+![Application form: select Fireworks AI under Product Needed](https://github.com/user-attachments/assets/12a49136-0956-4f29-9d47-f0e473615075)
+
+Complete the other required fields marked with `*`, double-check your email, identity, product selection, and profile link, then submit.
+
+## Step 4: Wait for the review
+
+AMD verifies your account and application, usually within **2–3 business days** — actual time varies with application volume, information completeness, and holidays.
+
+## Step 5: Receive and redeem the coupon code
+
+Once approved, AMD emails a unique **Coupon Code** redeemable for **$50 in Fireworks AI Credits** to your application address. Keep it secure — do not share it publicly, forward it, or commit it to a repository.
+
+![The Coupon Code in the approval email](https://github.com/user-attachments/assets/c53129c7-2c87-4510-a4e3-31f52598dc24)
+
+Redeem it and create an API key:
+
+1. Open <https://fireworks.ai/> and sign in;
+2. Click **Redeem Promo** and enter the Coupon Code from the email to redeem the $50 credits;
+3. Click **Create API Key** to generate your Fireworks API key.
+
+![Redeeming and creating an API key in the Fireworks console](https://github.com/user-attachments/assets/051a1e69-db7f-4867-b899-89981df15142)
+
+## Set it up in PenguinHarness
+
+With the API key in hand, three steps:
+
+**1. Install and launch**
+
+```bash
+curl -fsSL https://penguin.ooo/install.sh | sh
+penguin web        # opens http://127.0.0.1:7364 (first login: admin / penguin-2026)
+```
+
+**2. Configure a Fireworks model**
+
+Open the Models page and find the **Fireworks AI** group, then use its bulk key button to paste the API key you just created. The group presets five models — GLM 5.2, Kimi K2.7 Code, DeepSeek V4 Pro, MiniMax M3, and DeepSeek V4 Flash — with base URLs and pricing pre-filled; set any of them as the default. You can also hit the group's speed-test button to measure real TTFT and TPS before choosing.
+
+**3. Start working**
+
+Head back to Chat and hand the Agent its first task — e.g. "Analyze data.csv and summarize quarterly sales".
+
+## References
+
+- [AMD AI Developer Program](https://developer.amd.com/ai-developer-program/)
+- [Official AMD Cloud Credits application video tutorial](https://www.youtube.com/watch?v=masSW53JkTY)
+- Application steps and screenshots are adapted from WhatGhost's guides ([中文](https://github.com/WhatGhost/whatghost_Notebooks/blob/main/other/AMD_AI_Developer_Program_Credits_%E7%94%B3%E8%AF%B7%E6%8C%87%E5%8D%97.md) / [English](https://github.com/WhatGhost/whatghost_Notebooks/blob/main/other/AMD_AI_Developer_Program_Credits_Application_Guide_EN.md)) — thanks to the original author
+- [PenguinHarness model configuration docs](https://penguin.ooo/docs/models)
@@ -0,0 +1,84 @@
+---
+title: 携手 AMD 开发者计划：免费领取 $50 Fireworks 额度，直连 PenguinHarness
+date: 2026-07-20
+category: news
+excerpt: 我们很高兴与 AMD AI Developer Program 合作，为大家带来 Fireworks 的免费兑换码——按本文申请 $50 Credits，再在 PenguinHarness 里三步用起来。
+---
+
+我们很高兴与 **AMD AI Developer Program**（AMD 开发者计划）合作，为大家带来 Fireworks 的免费兑换码：加入计划并通过审核，即可获得可兑换 **$50 Fireworks AI Credits** 的 Coupon Code。PenguinHarness 内置 Fireworks AI 网关分组——OpenAI 协议、预置 base URL 与五个模型，额度到手即刻可用。
+
+> 页面内容、Credits 金额与有效期可能调整，请以申请时的页面及审批邮件为准。
+
+## 第一步：加入 AMD AI Developer Program
+
+访问官方入口：<https://developer.amd.com/ai-developer-program/>，在 **Join the AMD AI Developer Program** 区域：
+
+- 没有 AMD ADP 账户：点击 **Create Account**，填写个人信息创建账户；
+- 已有账户：点击 **Log In**，按页面提示登录。
+
+![AMD AI Developer Program 加入页面](https://github.com/user-attachments/assets/47a3055b-9a95-40a1-80c6-e3bac7a9ac49)
+
+## 第二步：在 Member Perks 申请 Cloud Credits
+
+注册并登录后：
+
+1. 点击顶部导航栏中的 **Member Perks**；
+2. 找到 **Cloud Credit Options**；
+3. 点击底部的 **Request Cloud Credits**。
+
+![Member Perks 中的 Cloud Credit Options](https://github.com/user-attachments/assets/773f8cf1-d72f-4aa3-83b8-f31fc7c9ed9e)
+
+## 第三步：填写申请表
+
+进入申请表后填写个人信息：
+
+- **Product Needed** 处选择 **Fireworks AI**；
+- **Profile** 处提供至少一个公开资料用于账户验证：LinkedIn、GitHub、Portfolio、公司 / 学校主页、WeChat、WeCom 等均可。
+
+![申请表：Product Needed 选择 Fireworks AI](https://github.com/user-attachments/assets/12a49136-0956-4f29-9d47-f0e473615075)
+
+填完页面中其他带 `*` 的必填项，检查邮箱、身份、产品选项与公开资料链接无误后提交。
+
+## 第四步：等待审核
+
+AMD 会验证账户与申请资料，通常需要 **2–3 个工作日**；实际时间可能因申请量、资料完整度或节假日而变化。
+
+## 第五步：收到兑换码并兑换
+
+审核通过后，AMD 会向申请邮箱发送包含唯一 **Coupon Code** 的邮件，可兑换 **$50 Fireworks AI Credits**。请妥善保存，不要公开、转发或提交到代码仓库。
+
+![审批邮件中的 Coupon Code](https://github.com/user-attachments/assets/c53129c7-2c87-4510-a4e3-31f52598dc24)
+
+兑换与创建 API key：
+
+1. 打开 <https://fireworks.ai/> 并登录；
+2. 点击 **Redeem Promo**，输入邮件中的 Coupon Code，兑换 $50 Credits；
+3. 点击 **Create API Key**，生成 Fireworks API key。
+
+![在 Fireworks 控制台兑换并创建 API key](https://github.com/user-attachments/assets/051a1e69-db7f-4867-b899-89981df15142)
+
+## 在 PenguinHarness 中用起来
+
+拿到 API key 后，三步接入：
+
+**1. 安装并启动**
+
+```bash
+curl -fsSL https://penguin.ooo/install.sh | sh
+penguin web        # 打开 http://127.0.0.1:7364（首次登录：admin / penguin-2026）
+```
+
+**2. 配置 Fireworks 模型**
+
+进入「模型仓库」页，找到 **Fireworks AI** 分组，点击「统一配置 key」粘贴刚创建的 API key。分组预置了五个模型——GLM 5.2、Kimi K2.7 Code、DeepSeek V4 Pro、MiniMax M3、DeepSeek V4 Flash——base URL 与价格已填好，任选一个设为默认即可；也可以点组头的「测速」，实测各模型的 TTFT 与 TPS 再决定。
+
+**3. 开始使用**
+
+回到对话页，把第一个任务交给 Agent——例如「分析 data.csv，输出各季度销售额汇总」。
+
+## 参考链接
+
+- [AMD AI Developer Program](https://developer.amd.com/ai-developer-program/)
+- [AMD 官方 Cloud Credits 申请视频演示](https://www.youtube.com/watch?v=masSW53JkTY)
+- 申请步骤与截图整理自 WhatGhost 的申请指南（[中文](https://github.com/WhatGhost/whatghost_Notebooks/blob/main/other/AMD_AI_Developer_Program_Credits_%E7%94%B3%E8%AF%B7%E6%8C%87%E5%8D%97.md) / [English](https://github.com/WhatGhost/whatghost_Notebooks/blob/main/other/AMD_AI_Developer_Program_Credits_Application_Guide_EN.md)），感谢原作者
+- [PenguinHarness 模型配置文档](https://penguin.ooo/docs/models)
@@ -2,70 +2,114 @@
 title: "Introducing PenguinHarness: agents that build agents"
 date: 2026-07-17
 category: news
-excerpt: The first open-source harness with recursive self-improvement is here — lightweight, efficient and secure infrastructure covering everything from automatic agent construction to continuous self-evolution.
+excerpt: We proved agents can self-evolve in our GDPevo Benchmark — now we are bringing that capability to everyone. The first open-source harness with recursive self-improvement covers everything from one-sentence agent construction to continuous self-evolution.
 ---

-Today we are releasing **PenguinHarness** — an open-source harness built for constructing and evolving agents. Its purpose fits in one line:
+Today we are releasing **PenguinHarness** — an open-source harness built for constructing and evolving agents: a zero-code Harness CLI and Web UI, connected to 1000+ models. The story it tells fits in one line:

-> Efficient Self-Improving Harness for Everyone.
+> With LangChain, you build agents by hand — at 1× speed. With PenguinHarness, agents build agents — at 100×.
+
+## From GDPevo to PenguinHarness: why we built this
+
+Before PenguinHarness, our team published the [GDPevo Benchmark](https://prism-shadow.github.io/GDPevo/). In GDPevo we systematically verified one thing: **agents can self-evolve** — an Agent can score its own performance, find where the points were lost, rewrite its own prompts and Skills, and climb version after version.
+
+With the capability proven, the question became: how does everyone get to use it? Self-evolution should not stay a curve in a paper — it should be infrastructure that works out of the box on every developer's desk. **Bringing an efficient self-improving harness to everyone is why we built PenguinHarness** — and it is right there in the name: Efficient Self-Improving Harness for Everyone.

 ## Why PenguinHarness

-Over the past year the way agent applications are built has been converging fast: what really decides quality is not a heavyweight framework but a simple, reliable, observable harness. PenguinHarness is rebuilt from the ground up — no dependency on any agent framework, a fully open-source self-developed kernel — and it brings three things to the open-source world first:
+Three reasons, in deliberate order — from task quality, to how agents get built, to how they keep improving.

- **Simplest Is the Best**: a deliberately minimal toolset over clean low-level interfaces — fewer tool calls, fewer Tokens, complex tasks done efficiently.
- **Harness for Building Agents**: with the PenguinHarness SDK, an Agent builds complete Agent applications for you, autonomously, from scratch.
- **Harness for Recursive Self-Improvement**: with PenguinHarness Skills, an Agent evaluates and optimizes itself, improving recursively over time.
+### 1. Better on complex tasks, at lower cost

-For the latter two, PenguinHarness is the first open-source implementation in the industry.
+A deliberately minimal toolset over clean low-level interfaces: fewer tool calls, fewer Tokens, deeply tuned for open models like DeepSeek. Each harness runs the model it is normally paired with — the comparison is between the products as people actually use them — head-to-head on two suites:

-## Same model, equal or better quality, lower cost
+![Benchmark: PenguinHarness leads the data-analysis suite and ties OpenAI Codex on coding, at a small fraction of both rivals' cost](/blog-assets/benchmark-light.svg)

-All runs use the same DeepSeek V4 Pro model, head-to-head against Claude Code and OpenAI Codex on two suites (per-run means below).
+Complex data analysis (15 tasks, single run; PenguinHarness and Codex at thinking xhigh, Claude Code at max):

-Complex data analysis (15 tasks, single run):
+| Framework      | Model           | Accuracy (%) | Tokens (M) | Cost ($) |
+| -------------- | --------------- | -----------: | ---------: | -------: |
+| PenguinHarness | DeepSeek V4 Pro |        66.67 |      18.04 |     0.55 |
+| Claude Code    | Claude Opus 4.8 |        53.33 |      22.20 |    38.48 |
+| OpenAI Codex   | GPT-5.5         |        53.33 |      13.72 |    19.41 |

-| Framework | Model | Accuracy (%) | Tokens (M) | Cost ($) |
-| --- | --- | ---: | ---: | ---: |
-| PenguinHarness | DeepSeek V4 Pro | 66.7 | 18.04 | 0.552 |
-| Claude Code | DeepSeek V4 Pro | 66.7 | 21.17 | 0.641 |
-| OpenAI Codex | DeepSeek V4 Pro | 46.7 | 13.36 | 0.427 |
+Coding tasks (40 tasks × 2 runs; accuracy is over all 80 outcomes):

-Coding tasks (40 tasks × 2 runs averaged, thinking high, 30 min per-case timeout, CNY pricing converted at $1 = ¥7):
+| Framework      | Model           | Accuracy (%) | Tokens (M) | Cost ($) |
+| -------------- | --------------- | -----------: | ---------: | -------: |
+| PenguinHarness | DeepSeek V4 Pro |        71.25 |     200.00 |     3.81 |
+| Claude Code    | Claude Opus 4.8 |        86.25 |     151.61 |   146.97 |
+| OpenAI Codex   | GPT-5.5         |        71.25 |     251.20 |   220.08 |

-| Framework | Model | Accuracy (%) | Tokens (M) | Cost ($) |
-| --- | --- | ---: | ---: | ---: |
-| PenguinHarness | DeepSeek V4 Pro | 50.00 | 2.10 | 0.041 |
-| Claude Code | DeepSeek V4 Pro | 48.75 | 2.00 | 0.048 |
-| OpenAI Codex | DeepSeek V4 Pro | 42.50 | 2.65 | 0.043 |
+Tokens and cost are suite totals, not per-run means. On data analysis we take the highest accuracy of the three — 66.67% against 53.33% for both — while spending 1/35 of what Codex spent and 1/70 of Claude Code's. On coding we tie Codex at 71.25% and trail Claude Code's 86.25%, but the whole suite cost us $3.81 against their $220.08 and $146.97: comparable work, one to two orders of magnitude apart on the bill.

-On the data-analysis suite PenguinHarness ties Claude Code on accuracy and clearly beats OpenAI Codex while using 14.8% fewer Tokens at 13.8% lower cost; on the coding suite it scores highest of the three at the lowest per-run cost.
+### 2. One sentence, and an Agent builds your Agent app
+
+Type one sentence, and an Agent builds the complete Agent application for you — scaffold, code, and run instructions, end to end:
+
+```text
+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:
+
+![The generated RAG app: a Claude Code docs expert answering with cited, clickable sources and example questions](/blog-assets/rag-app-en-light.webp)
+
+**And generating this entire RAG app burned just $0.02 (¥0.2) of tokens — on DeepSeek V4 Pro.**
+
+### 3. Self-evolution: it gets stronger with use
+
+With PenguinHarness Skills, an Agent evaluates and optimizes itself: the Optimizer orchestrates multiple Evaluators to score in parallel, uses the scores and run traces to find where points were lost, and upgrades the Agent from version N to N+1 — with a snapshot before every round, and every request replayable in the Trace view. A self-evolution demo video is coming soon.

 ## Evolution within bounds, security first

-The biggest worry about self-improvement is losing control. PenguinHarness answers with a contract — CONTRACT.md:
+The biggest worry about self-evolution is losing control. PenguinHarness answers with a contract (CONTRACT.md):

- Evolution is strictly confined to Workspace and Skills; the harness core security boundary is never modified.
- Tool calls run only after approval, and every approval is audited.
- Risky changes snapshot first — every step of evolution can be rolled back.
- Fully open source and locally deployable: data never leaves your machine, meeting enterprise security requirements.
+- Evolution is strictly confined to Workspace and Skills — the harness core security boundary is never modified;
+- Tool calls require approval first, and every approval leaves an audit record;
+- Risky changes are preceded by version snapshots, so any round of evolution can be rolled back;
+- Fully open source and locally deployed — data never leaves your machine, meeting enterprise data-security requirements.

-## Get started now
+## Supported models

-Install with one command (Linux / macOS, x64 / arm64, bundled Node runtime):
+| Model            | Providers                                                                        |
+| ---------------- | -------------------------------------------------------------------------------- |
+| DeepSeek V4      | DeepSeek, OpenRouter, Fireworks AI, SiliconFlow, Qwen Token Plan                 |
+| 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 (preview)                                                        |
+| GPT 5.5          | OpenAI, OpenRouter                                                               |
+| Gemini 3.5 Flash | Google Gemini, OpenRouter                                                        |
+| Claude Opus 4.8  | Anthropic, OpenRouter                                                            |
+
+Any OpenAI-protocol endpoint is supported: pick a preset above, or point a custom endpoint at any of the 1000+ online and local models.
+
+## How to use it
+
+Install with one command (Linux / macOS, x64 / arm64, bundled Node runtime), then launch the Web UI:

 ```bash
-curl -fsSL https://github.com/Prism-Shadow/penguin-harness/releases/latest/download/install.sh | sh
+curl -fsSL https://penguin.ooo/install.sh | sh
+penguin web        # opens http://127.0.0.1:7364 (first login: admin / penguin-2026)
 ```

-Configure a model (DeepSeek as an example), run your first task, or open the desktop-grade web interface with `penguin web`:
+Open the Models page, paste an API key under the DeepSeek or OpenRouter group and set it as default; then head back to Chat and hand the Agent its first task — e.g. "Analyze data.csv and summarize quarterly sales".

-```bash
-penguin config model add --model-id deepseek-v4-pro --api-key sk-your-key --set-default
-penguin run --approve allow-all --message "Analyze data.csv and summarize quarterly sales"
-penguin web
-```
+## What's next

-PenguinHarness supports 1000+ online and local models and multi-agent collaborative evolution, and runs on as little as a single CPU. Through continuous evolution it makes complex AI development ever simpler — a more efficient, more reliable, lower-hallucination and lower-cost Agent productivity engine.
+- Public release of the benchmark suite;
+- A desktop app;
+- Windows support;
+- More to come.

-Follow us on [GitHub](https://github.com/Prism-Shadow/penguin-harness) and open your first issue.
+## Join the community and build with us
+
+A self-improving harness needs a community that improves with it. Come discuss, request, and contribute — your first Issue is the best way to start:
+
+- [Discord](https://discord.gg/eFHKqqcU3D): chat with us and other developers in real time;
+- [X (Twitter)](https://x.com/code_hiyouga): follow the latest updates;
+- [WeChat group](https://github.com/Prism-Shadow/penguin-harness-community/blob/main/wechat/group.jpg): Chinese community discussions;
+- [GitHub](https://github.com/Prism-Shadow/penguin-harness): stars, Issues, and PRs all welcome.
+
+Self-evolving agent infrastructure, for everyone — starting today.
@@ -2,44 +2,64 @@
 title: PenguinHarness 正式发布：让 Agent 为你构建 Agent
 date: 2026-07-17
 category: news
-excerpt: 首个支持递归自我进化的开源 Harness 正式发布——以轻量、高效、安全的方式，提供从 Agent 自动构建到持续自我进化的完整基础设施。
+excerpt: 我们在 GDPevo Benchmark 中验证了 Agent 自我进化的能力，现在把它带给所有人——首个支持递归自我进化的开源 Harness 正式发布，从一句话构建 Agent 到持续自我进化，一套基础设施全部覆盖。
 ---

-今天，我们正式发布 **PenguinHarness**——一个为构建与进化 Agent 而生的开源 Harness。它的主旨只有一句话：
+今天，我们正式发布 **PenguinHarness**——一个为构建与进化 Agent 而生的开源 Harness：零代码的 Harness CLI 与 Web UI，连接 1000+ 模型。它要讲的故事只有一句话：

-> Efficient Self-Improving Harness for Everyone.
+> 使用 LangChain，以 1 倍速度人工构建 Agent；使用 PenguinHarness，以 100 倍速度用 Agent 构建 Agent。
+
+## 从 GDPevo 到 PenguinHarness：我们的初心
+
+在 PenguinHarness 之前，我们团队发布了 [GDPevo Benchmark](https://prism-shadow.github.io/GDPevo/)。在 GDPevo 中，我们系统性地验证了一件事：**Agent 可以自我进化**——让 Agent 评估自己的表现、定位失分原因、改写自己的提示词与技能，分数随版本一路上升。
+
+能力验证了，问题就变成了：怎么让每个人都用上它？自我进化不该只是论文里的曲线，而应该是每个开发者桌面上开箱即用的基础设施。**让每个人都能使用 Efficient Self-Improving Harness，这就是我们构建 PenguinHarness 的初心**——它也因此得名：Efficient Self-Improving Harness for Everyone.

 ## 为什么是 PenguinHarness

-过去一年里，Agent 应用的开发范式在快速收敛：真正决定效果的不是庞大的框架，而是一个简洁、可靠、可观测的 Harness。PenguinHarness 从底层重构，不依赖任何 Agent 框架，开源自研 Harness 内核，并率先把三件事带入开源世界：
+三个递进的理由——从任务效果，到构建方式，再到进化能力。

- **Simplest Is the Best**：坚持最小化工具集与简洁的底层接口，以更少的工具调用与 Token 消耗，高效完成复杂任务。
- **Harness for Building Agents**：通过 PenguinHarness SDK，让 Agent 从零自主完成 Agent 应用的构建。
- **Harness for Recursive Self-Improvement**：通过 PenguinHarness Skills，Agent 以自我评估与自我优化实现递归式自我提升。
+### 1. 复杂任务表现更好，成本更低

-后两项能力，PenguinHarness 是业内首个开源实现。
+刻意精简的工具集配合干净的底层接口：更少的工具调用、更少的 Token，对 DeepSeek 等开放模型深度适配。每个产品搭配它常用的模型——比的是大家实际会怎么用——在两套题库上正面对比：

-## 同一模型，同级效果，更低消耗
+![Benchmark：PenguinHarness 在数据分析题库准确率最高、编程题库与 OpenAI Codex 持平，成本仅为两者的零头](/blog-assets/benchmark-light.svg)

-全部使用同一 DeepSeek V4 Pro 模型，与 Claude Code、OpenAI Codex 在两套题库上正面对比（表中为单次运行均值）。
+复杂数据分析（15 题，单次运行；PenguinHarness 与 Codex 为 thinking xhigh，Claude Code 为 max）：

-复杂数据分析（15 题，单次运行）：
+| 实验框架       | 模型名称        | 准确率（%） | Token 用量（M） | 成本（$） |
+| -------------- | --------------- | ----------: | --------------: | --------: |
+| PenguinHarness | DeepSeek V4 Pro |       66.67 |           18.04 |      0.55 |
+| Claude Code    | Claude Opus 4.8 |       53.33 |           22.20 |     38.48 |
+| OpenAI Codex   | GPT-5.5         |       53.33 |           13.72 |     19.41 |

-| 实验框架 | 模型名称 | 准确率（%） | Token 用量（M） | 成本（$） |
-| --- | --- | ---: | ---: | ---: |
-| PenguinHarness | DeepSeek V4 Pro | 66.7 | 18.04 | 0.552 |
-| Claude Code | DeepSeek V4 Pro | 66.7 | 21.17 | 0.641 |
-| OpenAI Codex | DeepSeek V4 Pro | 46.7 | 13.36 | 0.427 |
+代码任务（40 题 × 2 次运行，准确率取全部 80 次结果）：

-代码任务（40 题 × 2 runs 取均值，thinking high、单题 30 分钟超时，人民币计价按 $1 = ¥7 折算）：
+| 实验框架       | 模型名称        | 准确率（%） | Token 用量（M） | 成本（$） |
+| -------------- | --------------- | ----------: | --------------: | --------: |
+| PenguinHarness | DeepSeek V4 Pro |       71.25 |          200.00 |      3.81 |
+| Claude Code    | Claude Opus 4.8 |       86.25 |          151.61 |    146.97 |
+| OpenAI Codex   | GPT-5.5         |       71.25 |          251.20 |    220.08 |

-| 实验框架 | 模型名称 | 准确率（%） | Token 用量（M） | 成本（$） |
-| --- | --- | ---: | ---: | ---: |
-| PenguinHarness | DeepSeek V4 Pro | 50.00 | 2.10 | 0.041 |
-| Claude Code | DeepSeek V4 Pro | 48.75 | 2.00 | 0.048 |
-| OpenAI Codex | DeepSeek V4 Pro | 42.50 | 2.65 | 0.043 |
+Token 与成本均为全套题目合计，不是单次均值。数据分析套件我们准确率最高——66.67% 对另两者的 53.33%——花的钱是 Codex 的 1/35、Claude Code 的 1/70；代码套件与 Codex 同为 71.25%、低于 Claude Code 的 86.25%，但整套题目只花了 $3.81，对方分别是 $220.08 与 $146.97：活干得差不多，账单差出一到两个数量级。

-数据分析套件与 Claude Code 准确率持平、显著超过 OpenAI Codex，同时 Token 消耗少 14.8%、成本低 13.8%；代码套件三者中准确率最高、单次成本最低。
+### 2. 一句话，让 Agent 构建 Agent 应用
+
+输入一句话，Agent 为你构建完整的 Agent 应用——脚手架、代码、运行说明，一步到位：
+
+```text
+收集 https://github.com/ericbuess/claude-code-docs 的文档，做一个化身 Claude Code 配置专家、回答带来源引用的 RAG 问答应用。
+```
+
+这是做出来的成品——一个文档专家：检索增强、引用可点击直达原文、内置示例问题：
+
+![生成的 RAG 应用成品：Claude Code 配置专家，回答带可点击的来源引用与示例问题](/blog-assets/rag-app-zh-light.webp)
+
+**而生成整个 RAG 应用，仅消耗了 0.2 元（$0.02）的 token——使用 DeepSeek V4 Pro 模型。**
+
+### 3. 自进化，越用越强
+
+借助 PenguinHarness 技能库，Agent 自己评估、自己优化：Optimizer 组织多个 Evaluator 并行打分，依据分数与运行轨迹定位失分原因，把 Agent 从版本 N 优化到版本 N+1——每轮之前自动快照，每个请求都可在轨迹观测中回放。自进化演示视频即将上线。

 ## 进化有界，安全先行

@@ -50,22 +70,46 @@ excerpt: 首个支持递归自我进化的开源 Harness 正式发布——以
 - 风险修改之前先留版本快照，任何一次进化都可回退；
 - 完全开源、本地部署，数据不出域，满足企业级数据安全。

-## 现在就可以开始
+## 支持的模型

-一行命令安装（Linux / macOS，x64 / arm64，内嵌 Node 运行时）：
+| 模型             | 可用供应商                                                                       |
+| ---------------- | -------------------------------------------------------------------------------- |
+| DeepSeek V4      | DeepSeek, OpenRouter, Fireworks AI, SiliconFlow, Qwen Token Plan                 |
+| 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（预览）                                                          |
+| GPT 5.5          | OpenAI, OpenRouter                                                               |
+| Gemini 3.5 Flash | Google Gemini, OpenRouter                                                        |
+| Claude Opus 4.8  | Anthropic, OpenRouter                                                            |
+
+只要是 OpenAI 协议的端点都可以接入：从上表选择预置，或用自定义端点连接 1000+ 在线与本地模型。
+
+## 如何使用
+
+一行命令安装（Linux / macOS，x64 / arm64，内嵌 Node 运行时），然后启动 Web 界面：

 ```bash
-curl -fsSL https://github.com/Prism-Shadow/penguin-harness/releases/latest/download/install.sh | sh
+curl -fsSL https://penguin.ooo/install.sh | sh
+penguin web        # 打开 http://127.0.0.1:7364（首次登录：admin / penguin-2026）
 ```

-配置模型（以 DeepSeek 为例）后即可运行第一个任务，或用 `penguin web` 打开桌面级 Web 界面：
+进入「模型仓库」页，在 DeepSeek 或 OpenRouter 分组里粘贴 API key 并设为默认；回到对话页，把第一个任务交给 Agent——例如「分析 data.csv，输出各季度销售额汇总」。

-```bash
-penguin config model add --model-id deepseek-v4-pro --api-key sk-your-key --set-default
-penguin run --approve allow-all --message "分析 data.csv，输出各季度销售额汇总"
-penguin web
-```
+## 发展计划

-PenguinHarness 支持 1000 多种在线与本地模型、多智能体协作进化，最低单 CPU 即可运行。通过不断进化，它会让复杂的 AI 开发越来越简单——为你提供更高效、更可靠、更低幻觉、更低成本的 Agent 生产力引擎。
+- Benchmark 套件正式发布；
+- 推出桌面端（Desktop）应用；
+- 支持 Windows 系统；
+- 更多规划，敬请期待。

-欢迎在 [GitHub](https://github.com/Prism-Shadow/penguin-harness) 上关注我们，提出你的第一个 Issue。
+## 加入社区，一起共建
+
+自我进化的 Harness，也需要一个不断进化的社区。欢迎加入讨论、提出需求、贡献代码——你的第一个 Issue 就是最好的开始：
+
+- [Discord](https://discord.gg/eFHKqqcU3D)：与我们和其他开发者实时交流；
+- [X（Twitter）](https://x.com/code_hiyouga)：关注最新动态；
+- [微信群](https://github.com/Prism-Shadow/penguin-harness-community/blob/main/wechat/group.jpg)：中文社区讨论；
+- [GitHub](https://github.com/Prism-Shadow/penguin-harness)：Star、Issue 与 PR 都欢迎。
+
+让每个人都用上会自我进化的 Agent 基础设施——从今天开始。