中文核心文件管理

PassAudited by ClawScan on May 8, 2026.

Overview

This instruction-only skill is coherent for managing OpenClaw core Markdown files, but it can read and edit persistent profile, memory, tool, and startup files in the workspace.

Install this only if you want an agent to manage OpenClaw core workspace files. Before allowing changes, ask it to show what it will edit, confirm backups, review diffs for user.md, memory.md, tools.md, agents.md, and bootstrap.md, and avoid keeping secrets in these Markdown files.

Findings (3)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Note

ASI02: Tool Misuse and Exploitation

What this means

The agent may inspect, back up, and edit local OpenClaw workspace Markdown files.

Why it was flagged

The skill explicitly requires read and write access to core files and optionally uses exec for validation and backup commands. This is aligned with file management, but it is still local mutation authority.

Skill content

| `read` | 读取核心文件 | 必需 | ... | `write` | 修改核心文件 | 必需 | ... | `exec` | 验证文件状态、备份 | 可选 |

Recommendation

Use it only for intended core-file maintenance, ask to see planned changes or diffs, and keep backups before writes.

Note

ASI06: Memory and Context Poisoning

What this means

Incorrect or unwanted edits could change what the agent remembers about the user or future workflows.

Why it was flagged

The skill manages files that hold user context and long-term memory. Edits to these files can persist and influence later agent behavior or expose private workspace details if mishandled.

Skill content

| 4 | **user.md** | 用户上下文 | 用户信息、偏好、基础设施 | ... | 5 | **memory.md** | 长期记忆 | 重要事件、关键配置、工作流程 |

Recommendation

Avoid storing secrets in these files, review memory/user changes carefully, and keep a restorable backup.

Note

ASI10: Rogue Agents

What this means

A bad edit to agents.md or bootstrap.md could make future OpenClaw sessions behave differently than expected.

Why it was flagged

The skill can manage persistent files that define agent operating rules and startup sequence. This is disclosed and purpose-aligned, but changes can affect future agent behavior.

Skill content

| 3 | **agents.md** | 逻辑与程序手册 | 规格、工作区规则、心跳、红线 | ... | **bootstrap.md** | 机器启动序列 | 不是灵魂 — 仅机器启动顺序 |

Recommendation

Require explicit review for changes to agents.md and bootstrap.md, and confirm backups exist before applying edits.