Security audit
Security checks for vulnerabilities and agentic risk
This is a note-taking workflow, but it can reorganize and modify an existing knowledge base and force follow-up audits, so users should review its scope before use.
Install only if you want this skill to manage a real knowledge vault, not just summarize text. Use it in a bounded vault or review proposed changes before allowing edits to existing indexes, file moves into 03_索引, companion skill calls, and workflow-audit remediation writes.
The skill requires metadata keys in English and specifically mandates that every tag value be in English, explicitly prohibiting Chinese tags. This is a natural-language locale policy constraint applied globally, and the file does not offer user opt-in or explain a region/compliance need that would justify forcing a specific language.
The skill’s stated purpose is deep reading and note generation, but it also mandates file placement, index onboarding, moving files, and modifying existing indexes across the repository. This expands its write scope from generating task-local outputs to altering unrelated knowledge-base structure, increasing the chance of unintended or user-unapproved changes.
The skill requires invoking additional skills and generating a mandatory workflow-audit report, which exceeds the declared scope of deep reading. Chaining compulsory downstream actions increases privilege and behavioral scope, and can cause extra file creation/modification unrelated to the user’s immediate request.
A mandatory process-auditing phase is not necessary to accomplish deep reading, yet it forces inspection of the task directory and creation of additional review artifacts until all checks pass. This creates pressure for repeated autonomous modifications and broadens operational authority beyond summarization or note synthesis.
The skill explicitly references another internal skill path, revealing environment structure and encouraging cross-skill dependency. Even if the disclosure is limited, enumerating internal capability locations can aid chaining behavior and expands knowledge of the local skill inventory beyond what is needed for the user task.
**Phase 6 完成后必须执行**。调用 **workflow-audit** skill，以德明+葛文德视角对本流程执行完成度逐项核对与系统闭环检查：
1. **审查对象**：本 skill（deep-reading）；**产出**：本次任务目录（执行计划、结构笔记、索引、原子笔记、task.md 等）。
2. **产出**：落盘审查报告（`YYYYMMDD_[任务名]_流程审查报告_德明与葛文德视角.md`），含逐项清单、系统闭环、DoD 勾选、多索引挂载清单；对审查中标为 ❌ 的项**必须补执行**，直至 DoD 全部通过。
3. **workflow-audit 主文件**：`.cursor/skills/workflow-audit/SKILL.md`；报告模板：`workflow-audit/references/audit_report_template.md`。
> 不可跳过。未通过流程执行审查并补全漏项，则本流程视为未完成。
The document is entirely written in Chinese and presents the methodology as the default operating mode, with no indication that users may choose another language or locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the regional constraint is explicitly justified.
This markdown file presents all user-facing instructions in Chinese and does not indicate that the user can select another language or that the skill is intentionally restricted to a Chinese-speaking context. That creates a natural-language policy concern under the language/locale rule because the skill implicitly forces a specific language without opt-in.
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
> 这是本 skill 的核心创新：**边创建边发现**。
**与主 SKILL 的衔接**：**显式链接全集**（从结构笔记提取、与 `task.md` 对齐）的定义与 **Phase 2.75 门禁**见 `.cursor/skills/deep-reading/SKILL.md` 的 **Phase 2.75**；**Phase 3 工作波次（Sweep 1/2/3+）**见同文件 **Phase 3**。本文只定义 **Scan 本身**及 **task.md 记录格式**。
---
The template includes fixed tag placeholders english-topic-tag and english-method-or-domain, which instruct users toward English-only metadata. This is a natural-language locale constraint, and the file does not offer an opt-in choice or explain why English is required.
The template contains Chinese prose throughout, but the metadata explicitly requires English tags via placeholders like english-topic-tag and english-subtopic-or-material. This creates a natural-language locale constraint without any user opt-in or explanation, which can violate language/locale policy expectations.
The template hard-codes English-only tag placeholders ("english-topic-tag", "english-subtopic-or-source") while the surrounding instructions and content are in Chinese. This imposes a specific language convention in the template without offering a language choice or documenting why that locale constraint is required.
The template headings and instructions are written entirely in Chinese, including file naming guidance and section labels, which implicitly requires use of that language. There is no indication that this is an opt-in locale-specific template or that users may choose another language.
No suspicious patterns detected.