How do I install the Openclaw Prime skill?

Run npx skillmds@latest add nicholasspisak/openclaw-prime in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Openclaw Prime safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: makes network calls, reads secrets. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Openclaw Prime?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Openclaw Prime free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Openclaw Prime?

nicholasspisak (@nicholasspisak) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Openclaw Prime by nicholasspisak — Agent Skill

Name: Openclaw Prime
Rating: 4.5 (4 reviews)
Author: nicholasspisak

Openclaw Prime

Prime a context window with OpenClaw infrastructure, gateway, channel, security, and operator knowledge from the OpenClaw docs. Use before answering admin questions, planning or executing configuration changes, debugging gateways or channels, reviewing deployments, or making operational decisions on an OpenClaw setup.

File contents

OpenClaw Prime

Overview

Load the OpenClaw docs into working memory before doing operator work. Use this skill to build a current mental model of the gateway, config, security model, channels, nodes, and deployment surfaces, then carry that context into the actual admin task.

Prime Workflow

1. Refresh the docs index

Start from the live docs, not memory.

Read https://docs.openclaw.ai/llms.txt to get the current page list.
If recency matters, check https://docs.openclaw.ai/sitemap.xml for recently changed pages.
Read admin-primer.md for the distilled operator map before diving into page-specific docs.

2. Read the core pages every time

Read these pages for almost any infra or admin task:

https://docs.openclaw.ai/index.md
https://docs.openclaw.ai/start/hubs.md
https://docs.openclaw.ai/gateway/index.md
https://docs.openclaw.ai/gateway/configuration.md
https://docs.openclaw.ai/gateway/configuration-reference.md
https://docs.openclaw.ai/gateway/security/index.md
https://docs.openclaw.ai/gateway/remote.md
https://docs.openclaw.ai/gateway/troubleshooting.md
https://docs.openclaw.ai/cli/index.md
https://docs.openclaw.ai/channels/index.md
https://docs.openclaw.ai/concepts/architecture.md
https://docs.openclaw.ai/concepts/multi-agent.md

3. Expand by task

Pick the smallest relevant doc set for the task.

Gateway auth or networking: gateway/authentication.md, gateway/tailscale.md, gateway/trusted-proxy-auth.md, gateway/network-model.md
Config or secrets: gateway/secrets.md, gateway/configuration-examples.md, auth-credential-semantics.md
Health or debugging: gateway/doctor.md, gateway/health.md, gateway/logging.md, help/troubleshooting.md
Channels or pairing: channels/channel-routing.md, channels/pairing.md, channels/troubleshooting.md, plus the specific channel page
Nodes or device surfaces: nodes/index.md, the relevant node page, and platform docs
Deployment or upgrades: install/index.md, the relevant install target page, install/updating.md, install/migrating.md
Model behavior: concepts/model-providers.md, concepts/model-failover.md, concepts/models.md
Automations: automation/*.md pages relevant to the hook, cron, poll, or webhook workflow

4. Separate docs truth from local truth

After priming from docs, inspect the actual environment before recommending or executing changes.

Identify deployment target: local, VM, Docker, cloud, macOS bundle, or remote gateway.
Identify config source: ~/.openclaw/openclaw.json, env overrides, secret storage, reverse proxy, tailnet, service manager.
Identify active surfaces: channels, nodes, models, auth mode, exposed ports, paired devices.
Identify risk boundaries: bearer tokens, trusted proxies, allowlists, sandbox and elevated policy.

Do not assume the local setup matches the docs defaults.

5. Produce a short operator brief

Before acting, summarize:

What OpenClaw component is involved
What config or secret surface matters
What docs pages govern the change
What operational risks exist
What is still unknown from the live environment

Keep it short. Enough to unblock the real admin task.

Output Standard

When this skill is invoked, leave the user with:

A concise OpenClaw architecture and ops summary
The exact docs pages that matter for the next action
The local-environment facts still needed before making changes

If a task depends on changing docs-sensitive behavior, re-open the live page immediately before giving final guidance.

---name: openclaw-primedescription: Prime a context window with OpenClaw infrastructure, gateway, channel, security, and operator knowledge from the OpenClaw docs. Use before answering admin questions, planning or executing configuration changes, debugging gateways or channels, reviewing deployments, or making operational decisions on an OpenClaw setup.---# OpenClaw Prime## OverviewLoad the OpenClaw docs into working memory before doing operator work. Use this skill to build a current mental model of the gateway, config, security model, channels, nodes, and deployment surfaces, then carry that context into the actual admin task.## Prime Workflow### 1. Refresh the docs indexStart from the live docs, not memory.- Read `https://docs.openclaw.ai/llms.txt` to get the current page list.- If recency matters, check `https://docs.openclaw.ai/sitemap.xml` for recently changed pages.- Read [admin-primer.md](./references/admin-primer.md) for the distilled operator map before diving into page-specific docs.### 2. Read the core pages every timeRead these pages for almost any infra or admin task:- `https://docs.openclaw.ai/index.md`- `https://docs.openclaw.ai/start/hubs.md`- `https://docs.openclaw.ai/gateway/index.md`- `https://docs.openclaw.ai/gateway/configuration.md`- `https://docs.openclaw.ai/gateway/configuration-reference.md`- `https://docs.openclaw.ai/gateway/security/index.md`- `https://docs.openclaw.ai/gateway/remote.md`- `https://docs.openclaw.ai/gateway/troubleshooting.md`- `https://docs.openclaw.ai/cli/index.md`- `https://docs.openclaw.ai/channels/index.md`- `https://docs.openclaw.ai/concepts/architecture.md`- `https://docs.openclaw.ai/concepts/multi-agent.md`### 3. Expand by taskPick the smallest relevant doc set for the task.- Gateway auth or networking: `gateway/authentication.md`, `gateway/tailscale.md`, `gateway/trusted-proxy-auth.md`, `gateway/network-model.md`- Config or secrets: `gateway/secrets.md`, `gateway/configuration-examples.md`, `auth-credential-semantics.md`- Health or debugging: `gateway/doctor.md`, `gateway/health.md`, `gateway/logging.md`, `help/troubleshooting.md`- Channels or pairing: `channels/channel-routing.md`, `channels/pairing.md`, `channels/troubleshooting.md`, plus the specific channel page- Nodes or device surfaces: `nodes/index.md`, the relevant node page, and platform docs- Deployment or upgrades: `install/index.md`, the relevant install target page, `install/updating.md`, `install/migrating.md`- Model behavior: `concepts/model-providers.md`, `concepts/model-failover.md`, `concepts/models.md`- Automations: `automation/*.md` pages relevant to the hook, cron, poll, or webhook workflow### 4. Separate docs truth from local truthAfter priming from docs, inspect the actual environment before recommending or executing changes.- Identify deployment target: local, VM, Docker, cloud, macOS bundle, or remote gateway.- Identify config source: `~/.openclaw/openclaw.json`, env overrides, secret storage, reverse proxy, tailnet, service manager.- Identify active surfaces: channels, nodes, models, auth mode, exposed ports, paired devices.- Identify risk boundaries: bearer tokens, trusted proxies, allowlists, sandbox and elevated policy.Do not assume the local setup matches the docs defaults.### 5. Produce a short operator briefBefore acting, summarize:- What OpenClaw component is involved- What config or secret surface matters- What docs pages govern the change- What operational risks exist- What is still unknown from the live environmentKeep it short. Enough to unblock the real admin task.## Output StandardWhen this skill is invoked, leave the user with:1. A concise OpenClaw architecture and ops summary2. The exact docs pages that matter for the next action3. The local-environment facts still needed before making changesIf a task depends on changing docs-sensitive behavior, re-open the live page immediately before giving final guidance.

Frequently asked questions