Featured

Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger
Crawl and scrape any site into clean data, 10% off logoCrawl and scrape any site into clean data, 10% off

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits, and new users get 10% off their first purchase.

Try Firecrawl free
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free
SetupClaw: done-for-you OpenClaw for founders & exec teams logoSetupClaw: done-for-you OpenClaw for founders & exec teams

White-glove OpenClaw for founders and exec teams (4–50+ employees): we install, harden, integrate your tools, and maintain it — secured from day one.

Get it set up for you
SEO data APIs for your agent, $1 free credit logoSEO data APIs for your agent, $1 free credit

DataForSEO gives your agent live access to SERP results, keyword data, backlinks, and on-page SEO data through one API. New accounts get a $1 credit, good for up to 20,000 keyword or backlink lookups.

Try DataForSEO free
Reach 47,000+ AI builders

A flat monthly placement in front of developers actively installing AI tools. No lock-in, cancel anytime.

Advertise here
my-plugin logo

my-plugin

tbdavid2019/openclaw-docs-skill

Otheropenclawby tbdavid2019

Summary

OpenClaw plugin exposing 0 skills.

Install to Claude Code

openclaw plugin add tbdavid2019/openclaw-docs-skill

Run in Claude Code. Add the marketplace first with /plugin marketplace add tbdavid2019/openclaw-docs-skill if you haven't already.

README.md

OpenClaw Documentation Skill

繁體中文 | English

This repository turns the official OpenClaw documentation into an auto-updated Agent Skill. GitHub Actions mirrors the upstream docs/ directory daily, records the exact source revision, and generates compact topic catalogs for progressive disclosure.

What the skill helps with

  • Installation, updates, migrations, and deployments
  • Gateway configuration, operation, networking, and security
  • Messaging channels such as Telegram, WhatsApp, Discord, Slack, LINE, and Signal
  • Model providers, authentication, failover, and local models
  • Agents, sessions, memory, tools, plugins, nodes, and automation
  • CLI errors and symptom-first troubleshooting

Progressive disclosure

The full documentation set is intentionally not injected into the model context.

SKILL.md
  → compact references/SKILL_INDEX.md
  → one references/_catalog/<topic>.md
  → at most 1–3 relevant documents
  → only the relevant sections

Exact errors, CLI commands, and configuration keys bypass the catalog and are searched directly across references/. The generated catalogs use each upstream page's title, summary, and

read_when metadata.

Install

Install as a Git checkout so it can be updated reliably. Replace the destination with the skills directory used by your agent:

git clone https://github.com/tbdavid2019/openclaw-docs-skill.git \
  ~/.gemini/antigravity/skills/openclaw-docs

For Codex, a typical destination is:

git clone https://github.com/tbdavid2019/openclaw-docs-skill.git \
  ~/.codex/skills/openclaw-docs

After a direct clone, run the installer once to pull with fast-forward-only semantics, validate the snapshot, and print its commits:

bash <skill-directory>/scripts/install-skill.sh

The installer can instead clone into a new destination when run from another checkout:

bash openclaw-docs-skill/scripts/install-skill.sh <skill-directory>

To update an installed checkout later:

bash <skill-directory>/scripts/install-skill.sh

The public repository is synchronized daily. An installed checkout does not perform a network update on every question; it changes only when the installer or git pull is explicitly run.

If a host platform copies skills into a managed directory, use that platform's update mechanism. The copied snapshot remains usable but cannot update itself as a Git checkout.

AI Agent Installation Contract

When a user gives you this repository URL and asks you to install the skill:

1. Detect the host and resolve its actual skills directory before writing. For Codex, prefer

$CODEX_HOME/skills/openclaw-docs when CODEX_HOME is set; otherwise use

~/.codex/skills/openclaw-docs. A typical Antigravity destination is

~/.gemini/antigravity/skills/openclaw-docs. For another host, inspect its configuration or ask the user instead of guessing. 2. Inspect the destination. If it does not exist, clone this repository with the final directory name openclaw-docs. If it is already this repository's Git checkout, keep it. Do not overwrite a non-Git directory; explain the conflict and ask the user how to proceed. 3. You MUST run the installer before verification:

bash <skill-directory>/scripts/install-skill.sh. For an existing checkout, the installer MUST successfully complete git pull --ff-only; a pull or validation failure means the installation or update failed. Run the installer after a new clone as well so the same validation is applied. 4. Read the installed SKILL.md. Do not run scripts/sync-docs.sh or

scripts/generate_index.py as part of installation. They are maintainer tools; an installation must use the repository's committed snapshot. 5. Verify that the destination contains SKILL.md, agents/openai.yaml,

references/SOURCE.json, references/SKILL_INDEX.md, and references/_catalog/. 6. Read references/SOURCE.json, then report the absolute installation directory, the Skill repository commit, and the upstream documentation commit. The installer prints all three. 7. Tell the user to reload or restart the host if it discovers skills only at startup. Then test discovery with: Use $openclaw-docs to explain how to check OpenClaw Gateway health. 8. Do not claim installation success until pull, validation, commit reporting, and any required reload instruction have completed. Never print credentials or unrelated configuration values.

Copy-paste installation request:

Install the following repository as the OpenClaw documentation skill:
https://github.com/tbdavid2019/openclaw-docs-skill

Follow README.md's "AI Agent Installation Contract":
1. Clone it into your skills directory with the final directory name openclaw-docs.
2. Run scripts/install-skill.sh. An existing checkout must complete git pull --ff-only.
3. Read SKILL.md.
4. Verify references/SOURCE.json, references/SKILL_INDEX.md, and references/_catalog/.
5. Report the installation directory, Skill repository commit, and upstream documentation commit.

After installation, test it with:
Use $openclaw-docs and the locally synchronized official OpenClaw documentation to explain how
to check Gateway health. Cite the local document paths and distinguish documentation facts from
your inferences.

The GitHub repository is synchronized daily, but an installed checkout does not update itself.
Re-run the installer, which performs git pull --ff-only, before relying on the latest snapshot.

Repository structure

openclaw-docs-skill/
├── SKILL.md
├── agents/openai.yaml
├── scripts/
│   ├── install-skill.sh
│   ├── sync-docs.sh
│   └── generate_index.py
├── tests/test_repo.py
├── .github/workflows/auto-sync.yml
└── references/
    ├── SOURCE.json
    ├── SKILL_INDEX.md
    ├── _catalog/
    └── <official OpenClaw documentation>

references/SOURCE.json identifies the upstream repository, commit, commit date, and document count represented by the current snapshot.

Maintainer workflow

Refresh from the official repository:

sh scripts/sync-docs.sh

Validate the skill and generated artifacts:

python3 -m unittest -v tests/test_repo.py
python3 scripts/generate_index.py --check

The sync is staged and validated before references/ is replaced. A failed fetch, low document count, missing required page, or invalid generated index leaves the current references intact.

Do not hand-edit synchronized documents, references/SOURCE.json,

references/SKILL_INDEX.md, or references/_catalog/.

Documentation source

Content is synchronized from the official OpenClaw repository and OpenClaw documentation.

License

AGPL-3.0. OpenClaw documentation remains subject to its upstream licensing terms.

Acknowledgments

Inspired by win4r/OpenClaw-Skill.

Related plugins

Browse all →