dsh-discovery:DSH 社区插件的只读搜索与安全审查入口

dsh-discovery 是一款基于 MIT 协议的 DSH 社区插件只读搜索工具,旨在解决第三方插件安装前的筛选与安全审查问题。其核心功能包括浏览 GitHub `dsh-plugin` 话题下的插件,提供分类、场景筛选及中英同义词搜索。在安全层面,它执行静态规则扫描进行确定性预检,并生成报告供 LLM 在会话中深度审查代码、依赖及信誉,审查通过方可安装,高危风险默认拒绝。此外,它支持 Agent 调用搜索与预检工具、已安装插件的更新检查(含安全审查)及缓存机制。该工具本身不执行安装、更新或卸载操作,也不加载远程代码,仅作为只读入口。用户需注意,尽管有预检和 LLM 审查,仍无法穷尽所有风

Read More
dsh-auto-review:DSH 的默认开启安全审查插件

dsh-auto-review是DSH(DeepSeek Harness)的原生安全审查插件,旨在将“权限控制”与“内容审查”分离,防止因授予完全访问权限而跳过安全审查。该插件具备两大核心能力:一是在执行写、编辑、Bash等高风险动作前进行本地规则拦截,默认采取“失败关闭”策略,即未明确审查前拒绝执行;二是在变更交付时,由独立的只读子代理进行审查,若发现高危问题默认要求修复,且审查中断或结果不可解析时不会误报为“干净”。 插件提供默认开启且跨重启保留的会话级Auto Mode,用户可通过命令在当前会话临时关闭。它支持Web和cc-tui profile,需在DSH环境下安装,要求Node 18

Read More
dsh-plugin-security-review:给 DSH 动态插件安装加一道静态安全审查

`dsh-plugin-security-review` 是 ShanHaiFish 开发的 DSH 插件安全审查守卫,采用 MIT 许可。它针对通过 `cordis_define` 和 `cordis_run` 安装或运行的动态插件进行静态源码安全审查。该工具提供 `plugin_security_review` 和 `plugin_security_audit`两个核心功能,依据 `BLOCK`、`ASK`、`WARN`、`ALLOW` 四级策略处理插件操作。 其主要特性包括支持浏览器端的人工审批弹窗(同意、同意+白名单、拒绝),并能在人工批准后持久化代码指纹,后续相同指纹自动放行。此外,

Read More
dsh-tiered-approval:DeepSeek Harness 的分级自动审查插件

`dsh-tiered-approval`是专为DeepSeek Harness(DSH)设计的分级自动审查插件,旨在解决工具调用审批中“全确认”效率低与“全放权”风险高的两难困境。该插件采用MIT许可,通过三层裁决机制替代简单的二进制审批:首先利用静态规则安全网,直接拦截递归删除、格式化等不可逆高危操作;其次由LLM审查员判断调用是否符合用户意图,输出允许、拒绝或询问指令;最后对拿不准的调用进行人工兜底确认。插件支持配置`autoApproveAligned`以在意图高度对齐时免弹窗,以及`alignDeny`允许特定高危操作降级为人工确认。审查强度随DSH访问模式(如只读、工作区写)自动调

Read More
Differential Review: Trail of Bits formalizes PR security differential review into an executable workflow

Differential-review is an Agent Skill released by Trail of Bits, which conducts security-focused reviews on PRs, commits and diffs: it classifies tasks by risk rather than line count, uses git history to detect regressions, quantitatively calculates the blast radius, treats missing tests as a higher risk factor, and enforces the output of a Markdown report with line numbers and attack scenarios. This article checks the official SKILL.md, plugin README and documentation website, explaining what problems it solves, how to install it in Claude Code / Codex / the general Skill catalog, as well as three official usage scenarios and inapplicable scenarios.

Read More
OpenAI Official Security-Best-Practices: Performing Security Reviews of AI According to Languages and Frameworks

OpenAI has provided the `security-best-practices` Skill in the `curated` directory of the openai/skills repository. It includes 10 security specifications for common frameworks in Python, JavaScript/TypeScript and Go, and supports secure-by-default coding, passive inspection and structured security reporting. This article introduces its workflow, references library, Codex/Cursor installation methods and typical prompts, to help developers embed verifiable security review capabilities into AI-assisted programming.

Read More