Openpencil Setup logo

Openpencil Setup

Organization
ZSeven-W
openpencil-setup

Connect and diagnose the local Rust OpenPencil editor for the Codex OpenPencil plugin. Use for initial setup, missing tools, or a disconnected local canvas.

Overview

PublisherZSeven-W
Repositorycodex-openpencil
Skill nameopenpencil-setup
Stars
140
Forks
17
Bundled files
1
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by ZSeven-W on GitHub. Read the source before you install it.

Installation

Install the Openpencil Setup AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/ZSeven-W/codex-openpencil/tree/main/plugins/openpencil/skills/openpencil-setup
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/ZSeven-W/codex-openpencil.git /tmp/codex-openpencil
mkdir -p .claude/skills
cp -r /tmp/codex-openpencil/plugins/openpencil/skills/openpencil-setup .claude/skills/openpencil-setup
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Openpencil Setup in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Openpencil Setup on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Openpencil Setup is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Connect OpenPencil

The plugin opens a managed interactive Rust canvas or connects to an existing desktop editor. First canvas use automatically downloads the matching complete runtime from pinned public GitHub Release assets and validates its SHA-256, payload manifest, version and platform before configuring or starting it.

Interactive canvas

  1. In Codex, call canvas_open_browser, optionally with the user's existing absolute .op path, then open its panelUrl in the Codex browser panel (open_in_codex, browser target) or @Browser. Use an external browser only when the user explicitly requests one. The source is copied into a recovery draft. If only open_canvas is available, its returned panelUrl works the same way. Codex currently filters local HTTP iframe origins, so its MCP Apps entrypoint shows a browser launcher. Do not request weaker browser security or claim an embedded canvas was verified. Its button prepares the same canvas URL and, where host text messaging is supported, sends one opening request to the current chat through ui/message. Open the supplied exact URL with open_in_codex (browser target) or @Browser; preserve its theme/launch fragment so the panel can confirm editor readiness. Do not create a second canvas, launch an OS browser or use a deep-link workaround. Without host messaging, the user copies an opening instruction into the current chat. A message acknowledgment is not a connected canvas. ui/open-link acknowledgments also do not prove navigation: the current bridge drops HTTP URLs.
  2. Missing runtime configuration triggers installation automatically. Allow the initial download to finish; do not repeatedly call either opener after a host timeout. For preflight or a slow connection, run node scripts/runtime-install.mjs from the installed plugin directory, then call canvas_open_browser. Failed downloads do not configure partial runtimes. Existing invalid configurations are reported without switching runtimes. Configuration stays private in ~/.codex-openpencil/runtime.json. A complete custom package can be selected with OPENPENCIL_EDITOR_RUNTIME_DIR; it needs a matching native server, editable WASM, CanvasKit and integrity manifest. A desktop binary alone is insufficient.
  3. Use the UI resource/thread entrypoint, or show the returned local panelUrl in hosts without MCP Apps. Verify canvas_status and get_editor_state: they must describe the same managed document before editing.
  4. Each conversation keeps one canvas. Use another chat for a different document. Recovery drafts survive in ~/.codex-openpencil/sessions/. A stopped session's URL is temporary; reopen its recovery file in a new session.
  5. Reconnect or open a new chat after upgrading the plugin. Resource loading and visibly displayed embedded host UI are separate verification steps.

For a new install, follow the public repository's INSTALL.md. Its npm run setup verifies an isolated native edit/save/reopen before installing the plugin, registers the current Node executable privately, and verifies canvas tools through Codex's actual plugin loader. For a desktop startup error such as No such file or directory despite a working terminal, rerun setup and npm run host-check from that repository. Do not rewrite unrelated MCP settings. A successful fresh host check does not prove this existing chat has reloaded. If this chat has not loaded newly installed MCP tools yet, npm run canvas in the repository provides the same editable canvas and accepts native tool requests through a persistent terminal's stdin. Use that entry point to complete the design now, then explain that reconnect loads the installed tools. Never mistake a successful install for a verified live canvas.

Existing desktop editor

  1. Try get_editor_state through this plugin's OpenPencil tools. A response proves the actual connection; an installed package alone does not.
  2. If unavailable, run node scripts/mcp-bridge.mjs --doctor from the installed plugin directory (two levels above this file). The doctor reads the editor's ~/.openpencil/.op-mcp-port and verifies the instance identity. The plugin's launcher uses the Node.js 22+ executable registered by setup, then known local Node locations if needed; it never downloads a runtime to solve discovery.
  3. For a missing editor, use the official OpenPencil installation instructions. When launch is within the user's request, run op start, then enable MCP in the editor's Agent Settings. An old op binary may not discover the current Rust editor; trust the bridge doctor and editor state over that binary's status.
  4. For an intentional headless/web endpoint without a live discovery record, set OPENPENCIL_MCP_PORT in the MCP process environment or append --port PORT to the bundled server args. Only local loopback ports are accepted. Do not guess a port or rewrite unrelated Codex server configuration.
  5. If MCP discovery failed at session startup, reconnect/reload the plugin or start a new chat after the editor is reachable. Verify get_editor_state again and report the active document/page before editing.

A stale identity error requires reopening the intended editor or fixing its MCP setting. The bridge refuses to silently fall back to a different canvas. Keep discovery tokens out of chat, logs, committed files, and reports.

After either canvas opener, all native tools target that managed session instead of the discovered desktop editor. Report the target before edits; never silently take over a different document.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Openpencil Setup AI skill do?

Connect and diagnose the local Rust OpenPencil editor for the Codex OpenPencil plugin. Use for initial setup, missing tools, or a disconnected local canvas.

Why use Openpencil Setup on TypingMind?

Because you install it once and use it with any model. Openpencil Setup is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Openpencil Setup in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/ZSeven-W/codex-openpencil/tree/main/plugins/openpencil/skills/openpencil-setup. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Openpencil Setup?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Openpencil Setup?

As many as you like. As long as a model supports skills, you can use Openpencil Setup with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Openpencil Setup AI skill free?

Yes. It is published on GitHub by ZSeven-W under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇