Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese workflow-note writing template with some publication and example-memory cautions, but no hidden execution or malicious behavior was found.
Before using this skill to publish notes, review any copied cron prompts, AGENTS.md, HEARTBEAT.md, or similar configuration for secrets, personal data, internal addresses, and private instructions. Treat build/commit/push as actions that should require your explicit intent.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
## 整体架构
\`\`\`
交互发生
├─→ memory/YYYY-MM-DD.md 情形记忆（发生了什么）
└─→ .learnings/LEARNINGS.md 反思记忆（学到了什么）
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.
## 写作流程
1. **先读** `~/.openclaw/skills/note-taking/SKILL.md`（全局规范：命名、语言、侧边栏等）
2. **先查重**：搜索已有 workflows/ 笔记，同主题不重复创建
3. **先想主题**：提炼一个明确的技术主题，不是按时间堆砌做了什么
4. **按模板写**：使用下方模板
The instruction "所有正文内容使用中文" forces a specific language for all body content. This is a natural-language policy concern because the file does not offer the user a language choice or explain a justified region-specific requirement.
The rule "正文必须中文" bars English for the main content and imposes a fixed language policy. Because no opt-in, alternative, or justified compliance context is provided, this matches the language/locale policy violation category.
The example explicitly promotes persistent file-based memory and an automatic nightly review process that can copy, retain, and modify user-derived information over time. Without privacy, retention, consent, and scope warnings, users may unknowingly store sensitive data or allow automated changes to long-lived files, increasing confidentiality and integrity risk.
The setup instructions direct users to create persistent files under the home directory for long-term memory and error tracking, but do not warn that these files can accumulate secrets, personal data, or other sensitive operational context. Because they are persistent and centrally located, they become an attractive source of data leakage or unintended reuse.
The instructions create a persistent session-like memory store under the user's home directory, including files for learnings, errors, and feature requests. In an agent context, this enables durable accumulation of conversation-derived state across sessions, which can expose sensitive data, propagate stale or harmful context, and expand the blast radius of compromise.
安装后必须手动初始化：
\`\`\`bash
mkdir ~/.openclaw/workspace/.learnings
touch ~/.openclaw/workspace/.learnings/LEARNINGS.md
touch ~/.openclaw/workspace/.learnings/ERRORS.md
touch ~/.openclaw/workspace/.learnings/FEATURE_REQUESTS.md
The natural-language content appears to assume Chinese as the required language for the skill example, and there is no indication that users may choose another language or that the locale is intentionally restricted. This can be a language-policy issue under SQP-3 when a skill forces a specific language without opt-in.
No suspicious patterns detected.