Use Symbola in your AI tools
Your assistant can search packs, read what a pack decides, and give your coding agent the steps to build it. Every tool below reaches the same address. No account, no sign-in.
https://symbola.sh/mcp Chat assistants
- ChatGPT. Open Plugins, choose +, then Add custom MCP server. Name it Symbola, paste the address above, and choose no authentication. Your plan or workspace settings decide whether custom servers are allowed.
- Claude (web, desktop, mobile). Settings, Connectors, then find Symbola in the directory or choose Add custom connector and paste the address. More on Claude.
Coding agents and editors
- Cursor.
- VS Code (Copilot agent mode). Add to VS Code
$ code --add-mcp '{"name":"symbola","type":"http","url":"https://symbola.sh/mcp"}' - Claude Code.
$ claude mcp add --transport http symbola https://symbola.sh/mcp - Codex.
$ codex mcp add symbola --url https://symbola.sh/mcp - Gemini CLI.
$ gemini mcp add --transport http symbola https://symbola.sh/mcp - Anything else. Any client that speaks MCP over Streamable HTTP can use the same address.
The Symbola skill
A coding agent with the skill knows how to find a pack and build it with the symbola CLI,
without the connector. It works in Claude Code, Codex, Cursor, Copilot and other agents that read
SKILL.md skills.
$ npx skills add kausalflow/symbola --skill symbola-build In Claude Code you can install it as a plugin instead, which also adds the connector:
/plugin install symbola --marketplace kausalflow/symbola.
What you can ask
- "Find a pack for a research-paper library."
- "What does the Rondel pack decide about link keys?"
- "Show me Lightkone's PRODUCT.md."
- "Build kausalflow/homeostat in Go." A coding agent then builds it; a chat assistant gives you the request to paste into one.
The connector only reads the public registry. See Privacy and Terms.