PerryLink/dsh-auto-review 预览 preview

PerryLink/dsh-auto-review

DeepSeek Harness审批请求的Second-model AI自动审查:一个只读的审查子代理返回结构化的允许/拒绝判定及理由,默认失败关闭,且可从会话日志中完全审计(审批/询问 -> 自动审查/判定 -> 审批/决定)。

Project Overview项目介绍

This is a native plugin built exclusively for DeepSeek Harness, adding a second-model read-only reviewer subagent to DSH’s approval chain. It triggers automatically whenever an action crosses the sandbox boundary, then the subagent reads relevant evidence from the workspace to output a structured allow or deny verdict. It defaults to a fail-closed policy that blocks unsafe operations without requiring regular human input for routine approvals. You can install it via DSH Desktop’s built-in market by adding the PerryLink catalog source and confirms plugin identity via npm’s built-in verification before completing installation.

This plugin is designed for users who prioritize AI operation safety within DeepSeek Harness. Its typical workflow starts when the main agent requests a cross-boundary tool action, which automatically triggers the reviewer subagent. The subagent only uses allowed read-only tools to access workspace content, request parameters, and configured risk rules, then outputs a structured verdict with a clear reason for any denial. All decisions are logged as part of the session audit trail so you can reconstruct the full approval flow later and it supports configurable per-tool approval policies that you can adjust to your own risk tolerance.

dsh-auto-review requires Node.js version ^22.19.0 or >=24.0.0, and is compatible with all DeepSeek Harness versions from >=0.1.2-rc.1 to <0.2.0. There are no platform-specific limits, and it works on all operating systems that support DeepSeek Harness. It is released under the open-source Apache 2.0 license, with no hidden costs or paid tiers, which allows free use, modification, and redistribution for both personal and commercial projects. To install it, you first add the PerryLink catalog source to DSH Desktop’s built-in market, then select and confirm installation, which goes through npm identity verification before completion.

这是一个专为DeepSeek Harness开发的原生插件,可通过配置自定义不同工具的审批策略,在DSH的操作审批流程中加入了由第二个大模型担任的只读审核子代理。当工具操作越出沙箱安全边界时,该子代理会读取相关上下文证据后给出允许或拒绝的明确裁决,默认采用失败关闭策略,避免不安全操作通过的同时无需人类介入日常审批。

该插件适合关注DSH AI操作安全的用户使用,对安全要求较高的团队和个人开发者都适用,典型工作流程为:当主代理发起越界操作请求后,自动触发审核子代理,子代理仅使用只读工具读取工作区内容、请求参数和风险规则,输出结构化裁决,拒绝结论会同步返回给主代理并记录完整审计轨迹。

该插件要求Node版本为^22.19.0或>=24.0.0,兼容DSH版本>=0.1.2-rc.1 <0.2.0,采用Apache 2.0开源许可证,支持配置每工具的审批策略,可通过DSH桌面应用的内置市场添加来源后安装。

Pre-install check安装前体检Compatibility · Security兼容性 · 安全性 no risk signal found未发现风险信号
  • This site's static screen found no obvious risk signal (stars, license, activity, manifest)本站静态筛查没发现明显风险信号(星标、许可证、更新活跃度、清单完整度)
DSH walks through these 9 checksDSH 会逐条核对这 9 项

Compatibility兼容性

  • DSH, Node, OS and profile requirementsDSH 版本 / Node 版本 / 操作系统 / profile 是否满足要求
  • External dependencies and runtimes (Electron / Python / Docker, ...)外部依赖与运行时(Electron / Python / Docker 等)是否齐备
  • Conflicts with installed plugins: command names, skill / tool names, ports, duplicate MCP registration与已装插件是否冲突:命令名、skill / tool 重名、端口占用、重复 MCP 注册

Security安全性

  • Repo matches the facts registered here; archived or abandoned?仓库是否与页面登记一致,是否归档或长期停更
  • Safety of preinstall / install / postinstall and install.sh / setup.ps1preinstall / install / postinstall 与 install.sh、setup.ps1 是否安全
  • curl|bash, download-then-execute, obfuscation, unrelated domains → stop immediatelycurl|bash、下载即执行、混淆代码、无关域名 → 立刻停止
  • Typosquatting or unmaintained packages among the new dependencies新增依赖里有没有 typosquatting 或无人维护的包
  • Requested permissions vs. what the feature actually needs申请了哪些权限、是否超出功能所需(filesystem / network / shell / clipboard)
  • Any sudo / admin requirement, plus uninstall and rollback是否要求 sudo / 管理员权限,以及卸载与回滚方式

Anything uncertain must be marked unknown with a note on how to confirm it. This site's signal screen is a static snapshot, not a security audit.拿不准的必须标「未知」并说明要我怎么确认。本站的信号筛查是静态快照,不能替代安全审计。

Or use CLI install (for developers)或使用命令行安装(适合开发者)

CLI Install命令行安装

dsh plugin --profile web add dsh-auto-review

把 PerryLink/dsh-auto-review 加入你的 DSH 配置(web profile)即可启用。

READMEREADME

🤖 dsh-auto-review

Gitee dshfind OpenSSF Scorecard

Second-model AI approval for DeepSeek Harness — a read-only reviewer subagent decides allow/deny on the approval chain, fail-closed by default.

When an action crosses the sandbox boundary, a second model reads the evidence and returns a verdict with a reason — so humans approve nothing while nothing unsafe slips through.

Official repository. This is the only official repository of dsh-auto-review, maintained by PerryLink. Same-name repositories under other accounts are not affiliated.

License DSH plugin dsh-doctor DSH Market Node CI Version npm version npm downloads

English · 简体中文 · Español · Português · हिन्दी

Showing the opening section of the README — the full document lives in the repository以上为 README 开头摘要,完整文档在仓库内 · View the full README on GitHub →在 GitHub 查看完整 README →

← 上一个 Prev dsh-liang-skin 下一个 Next dsh-trading →