← 返回 Skills 市场
lueashes

Claw Wiki

作者 lueashes · GitHub ↗ · v0.1.2 · MIT-0
cross-platform ⚠ suspicious
1051
总下载
1
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install claw-wiki
功能描述
Reference and refresh a local OpenClaw documentation snapshot for product usage, deployment, configuration, CLI commands, channels, gateway behavior, tools,...
使用说明 (SKILL.md)

Claw Wiki

Use this skill to answer OpenClaw questions from local source docs in openclaw_docs/, not memory, and to refresh that snapshot when the user explicitly asks for an update.

Safety Contract

This skill is a documentation lookup and maintenance skill, not a general remote execution or credential-handling tool.

Allowed behavior:

  • Read the bundled local docs snapshot under openclaw_docs/
  • Read local metadata in docs.json, references/, and state/
  • Search the local snapshot with deterministic text/file queries
  • When the user explicitly asks for a refresh, sync docs from the upstream OpenClaw documentation source and rebuild local indexes

Hard limits:

  • Only use refresh mode when the user explicitly asks for freshness verification, update, or sync
  • Treat the upstream OpenClaw docs repository as read-only source material
  • Only sync documentation content and docs.json into this skill
  • Only write inside this skill directory, specifically openclaw_docs/, docs.json, and state/
  • Do not push commits, open pull requests, upload this skill, or publish local files elsewhere
  • Do not read local credential stores, shell history, SSH keys, .env files, or unrelated workspace files
  • Do not execute downloaded install scripts or run arbitrary shell fragments from documentation examples
  • Do not treat example tokens, passwords, API keys, or secrets shown in docs as real credentials

If a requested action exceeds those limits, refuse it and explain the boundary.

The maintained compatibility target is the English docs snapshot. Locale-prefixed navigation entries in docs.json may exist, but validation only guarantees the English document set is usable.

Files

  • ./openclaw_docs/: markdown source docs
  • ./docs.json: optional metadata/navigation mapping
  • ./state/source-lock.json: current snapshot source, ref, commit, and sync metadata
  • ./state/docs-manifest.json: generated manifest of the local docs snapshot
  • ./state/topic-index.json: generated keyword-to-path topic index
  • ./state/latest-diff.md: latest update diff summary
  • ./references/routing-cheatsheet.md: intent-to-path quick router for broad questions
  • ./references/query-examples.md: ready-to-use search query patterns by user intent
  • ./references/update-workflow.md: refresh workflow for syncing and validating the local snapshot
  • ./scripts/: deterministic sync, validate, build, and diff helpers

Workflow

  1. Determine whether the user wants documentation lookup or snapshot maintenance.
    • For normal product questions, answer from the current local snapshot.
    • For "latest", "refresh", "update docs", or "is this outdated?" requests, read state/source-lock.json first and use references/update-workflow.md.
  2. Identify intent and route to likely subdirectories first.
    • If intent is broad or mixed, read references/routing-cheatsheet.md first.
    • If wording is clear but query terms are weak, read references/query-examples.md.
  3. Search with fast text search (rg) under openclaw_docs/; avoid broad full-tree reads.
  4. Read full content of the top 1-3 most relevant files before answering.
  5. Synthesize a direct answer and include exact file paths used.
  6. When freshness matters, report the current snapshot commit/date from state/source-lock.json and do not imply the snapshot is current unless it was just refreshed.
  7. If the user explicitly requests a refresh, run the scripted flow:
    • python scripts/sync_docs.py ...
    • python scripts/validate_docs.py
    • python scripts/build_knowledge.py
    • python scripts/diff_docs.py
  8. If docs are ambiguous or missing, say so explicitly and state what was searched.

Maintenance Guardrails

  • Prefer a local OpenClaw checkout when available.
  • If remote sync is needed, use the pinned upstream OpenClaw docs source only.
  • Limit refresh work to documentation ingestion, validation, indexing, and diff generation.
  • Review generated state/latest-diff.md before claiming the refresh succeeded.
  • If sync input looks unrelated to OpenClaw docs, stop and report the mismatch instead of forcing the update.

Update Principles

  • Refresh this skill by reading upstream OpenClaw docs into the local snapshot only.
  • Allowed: local repo reads, git fetch, git clone, copying docs/ and docs.json into this skill, and rebuilding local indexes.
  • Not allowed: git push, PR creation, uploading this skill folder, or sending local skill contents back to GitHub.
  • Treat upstream refs such as main, tags, or commits as read-only sources, not write targets.
  • Prefer a local OpenClaw checkout when available; otherwise clone/fetch the upstream repo without modifying any remote.

Routing Hints

  • Getting started/install/platform setup: start/, install/, platforms/
  • Runtime model and behavior: concepts/, gateway/, tools/
  • CLI commands: cli/
  • Integrations/providers: providers/, channels/, plugins/, nodes/
  • Failures and debugging: help/, diagnostics/, gateway/troubleshooting.md, channels/troubleshooting.md
  • Security topics: security/, tools/exec-approvals.md, gateway/security/
  • Experimental or planning material: experiments/ (only use when user asks for proposals/plans)

Search and Read Strategy

  • Start with targeted queries, then widen only if needed.
  • Prefer searching filenames first, then file content.
  • Exclude merged or generated outputs when present.
  • For very large files, jump to relevant sections first, then expand for surrounding context.
  • Do not rely on a single snippet when explaining behavior, constraints, or defaults.
  • Prefer local snapshot sync over per-question remote browsing for repeatability.

Answer Quality Bar

  • Ground each key claim in files actually read.
  • Include paths in the response so users can verify quickly.
  • Separate documented behavior from inference.
  • Do not invent flags, defaults, commands, or config keys not found in docs.
  • For freshness-sensitive answers, include the snapshot version from state/source-lock.json.
安全使用建议
This skill appears to do exactly what it claims: answer from a checked-in OpenClaw docs snapshot and optionally refresh that snapshot using the included Python scripts. Before installing or running a refresh: - Inspect the sync script (scripts/sync_docs.py) to verify it only clones/fetches the intended upstream (or respects a pinned repo/ref) and does not accept arbitrary URLs from untrusted inputs. - Note that refresh requires network, git, and will write inside the skill folder (openclaw_docs/, state/, docs.json). Prefer installing the skill in a workspace-local skills directory, not a globally shared location, unless you trust the source. - The static scanner flagged potential prompt-injection and unicode-control-character patterns in SKILL.md. Although the skill includes an explicit Safety Contract, manually review SKILL.md and the scripts for any hidden/obfuscated directives before granting the agent permission to run an automated refresh. - Disable or explicitly approve any autonomous refresh actions (do not let the agent run the sync flow without your confirmation). If you want extra assurance, run the maintenance scripts yourself from a controlled environment (with a vetted upstream repo) rather than allowing the skill to do remote sync autonomously.
功能分析
Type: OpenClaw Skill Name: claw-wiki Version: 0.1.2 The claw-wiki skill is a documentation management tool designed to maintain a local, searchable snapshot of OpenClaw documentation. It includes a suite of Python scripts (scripts/sync_docs.py, scripts/build_knowledge.py, scripts/diff_docs.py) that synchronize content from the official OpenClaw GitHub repository and generate metadata for searching. The scripts implement robust safety measures, including a hardcoded repository whitelist, strict regex validation for Git refs, and path verification to prevent writes outside the skill's directory. The SKILL.md file provides the AI agent with a clear safety contract and hard limits, explicitly forbidding the execution of arbitrary code from documentation or the accessing of sensitive user data like credentials and SSH keys.
能力评估
Purpose & Capability
Name/description match the repo contents. The bundle contains a large local docs corpus and maintenance scripts (sync, validate, build, diff) which are appropriate for a documentation lookup + refresh skill. No unrelated environment variables, binaries, or config paths are requested.
Instruction Scope
SKILL.md restricts runtime behavior to reading the checked-in docs and running the included Python maintenance scripts only when the user explicitly requests a refresh. That is coherent with the purpose. Minor caveat: refresh operations involve git clone/fetch and network access; if invoked with arbitrary repo URLs (the scripts accept repo params in README examples) they could pull unexpected content. SKILL.md states a pinned upstream should be used, but the scripts themselves should be reviewed before running a remote sync to ensure they enforce the expected source.
Install Mechanism
No install spec — instruction-only skill loaded from workspace. All code is included in the bundle (Python scripts and docs). README notes Python 3.10+, git, and ripgrep (rg) as recommended tools; there are no external archives or remote installers in the install spec.
Credentials
The skill requires no environment variables or credentials. SKILL.md and SECURITY.md explicitly forbid credential discovery/exfiltration and treat example tokens in docs as inert. The only resources touched are files inside the skill directory and upstream docs when a refresh is requested.
Persistence & Privilege
Skill is not always-enabled and requests no elevated platform privileges. It writes only inside its own directory (openclaw_docs/, docs.json, state/) during refresh according to its guardrails. It does not declare any cross-skill or system-wide configuration modifications.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install claw-wiki
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /claw-wiki 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v0.1.2
- Added a SECURITY.md file to document security practices. - Introduced a "Safety Contract" section in SKILL.md outlining strict behavioral boundaries and explicit limitations for sync, execution, and data usage. - Clarified allowed/forbidden skill behaviors, emphasizing risk reduction and refusal of unsafe/irrelevant operations. - Added "Maintenance Guardrails" section to strengthen process safety around documentation refresh and validation. - No functional code changes are indicated; this is a documentation and security policy update.
v0.1.1
No functional or documentation changes in this release. - Version bump to 0.1.1 with no detected file or content updates.
v0.1.0
Initial release of claw-wiki skill. - Provides local reference and refresh workflow for OpenClaw documentation. - Answers usage, deployment, configuration, CLI, channels, gateways, troubleshooting, and doc freshness/versioning questions using only the current local docs snapshot. - Supports explicit snapshot refresh and syncing from upstream OpenClaw docs. - Details file structure, workflows, update principles, and quality standards for reliable and verifiable answers. - Answers include exact file paths for source traceability.
元数据
Slug claw-wiki
版本 0.1.2
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

Claw Wiki 是什么?

Reference and refresh a local OpenClaw documentation snapshot for product usage, deployment, configuration, CLI commands, channels, gateway behavior, tools,... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1051 次。

如何安装 Claw Wiki?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install claw-wiki」即可一键安装,无需额外配置。

Claw Wiki 是免费的吗?

是的,Claw Wiki 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Claw Wiki 支持哪些平台?

Claw Wiki 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Claw Wiki?

由 lueashes(@lueashes)开发并维护,当前版本 v0.1.2。

💬 留言讨论