dsh-auto-evolve
一个自我进化的 DeepSeek Harness 插件:观察代理的运行方式,通过 LLM 对其自身资产提出改进方案,在沙箱试运行代理中验证每个方案,只应用通过验证的变更——带版本化账本,出现性能回退时自动回滚。
- 形态
- Plugin 插件
- 安装
- npm 包安装
- 需要密钥
- 否
- 近月下载
- 2,129
- 最近推送
- 23 天前更新
项目说明
From README
dsh-auto-evolve A self-evolving plugin for DeepSeek Harness ( dsh ).
It observes how the agent runs, proposes improvements to its own assets via the LLM, validates each proposal inside a sandboxed trial agent, and applies only verified mutations — with a versioned ledger and automatic rollback on regression.中文 The idea.
Instead of shipping a static skill set, the plugin owns a small genome of evolvable assets (skills, tool-result post-processors, prompt sections, guard policies).
Every runtime signal (tool failures, repeated calls, request errors) is collected; when a threshold crosses, a proposal cycle drafts a mutation, a sandboxed agent replays the failing episode with and without the change, and the change is applied only if the trial shows measurable improvement.
Every mutation is versioned in a durable ledger and can be rolled back.How it works The evolution loop Layer Module Wha
Original description
A self-evolving plugin for DeepSeek Harness (dsh). It observes how the agent runs, proposes improvements to its own assets via the LLM, validates each proposal inside a sandboxed trial agent, and applies only verified mutations — with a versioned ledger and automatic rollback on regression.
安装方式
npm 包安装 · 安装无需配置
dsh plugin --profile web add dsh-auto-evolvedsh plugin --profile tui add dsh-auto-evolvegit clone https://github.com/lispking/dsh-auto-evolve.git安全能力
静态扫描源码得到的权限画像 —— 回答的是「它要动什么」,而不是「它好不好」
- 调用模型
未发现已知红线(明文传输、安装期执行代码、凭据外传等)。
能力与红线来自 awesome-dsh-plugin 的静态扫描结果(CC0 公有领域),本站未逐行复核源码; 它只覆盖能被静态识别的模式,不等于安全审计结论。
为什么是这个分数
近 23 天内有提交;README 给出了可直接复制的安装命令;装完即可用,无需额外配置;近月 npm 下载 2129 次,star 不多但真在被使用。
- 维护活跃
- 86
- 实用度
- 72
- 生态热度
- 67
- 上手便捷
- 100
- 信号质量
- 100
- 数据置信度
- 100%
相关插件
penguin-harness
Prism-Shadow/penguin-harness
Unified and Stable RSI Platform
evolver-dsh-plugin
EvoMap/evolver-dsh-plugin
越用越聪明的 harness,告别金鱼记忆。每条 prompt 前,Evolver 从本地经验和 EvoMap 网络召回一条验证过的策略;回合结束后,它凭测试结果、改动是否被保留、你的下一句反馈给策略打分,再回报网络按证据加权。每一次成功都变成下一次的捷径,每一次失败都变成下一次的避坑指南。
Proteus
proteus-evolve/Proteus
Self-evolution for any agent harness. Plug in. Evolve. Measure.