MCP Servers
Model Context Protocol servers and skills for operating the ibl.ai platform from your AI agent.
Overview
iblai/api exposes ibl.ai platform capabilities as tools an AI agent can call, enabling deep integration between language models and the platform. It ships two things: a set of skills that teach your agent to drive the platform REST API directly, and a hosted MCP server for the one runtime capability that is not a REST call — chatting with a deployed agent.
Where iblai/vibe gives you components to build an app, this gives you the means to run the platform itself: configure agents, manage datasets and memory, administer users and roles, send notifications, and pull analytics for any organization you belong to.
Each capability maps to one skill and one set of exact REST endpoints — method, URL, body — so changing an agent's LLM or pulling cost analytics is a single command rather than a documentation hunt.
Prerequisites
Node.js for npx, a skills-compatible AI agent (Claude Code, Cursor, OpenCode, and others), and an ibl.ai account with an organization. Signing up at ibl.ai/join creates both.
Repository
- GitHub: iblai/iblai-mcp
- License: MIT
Getting Started
Install the skills:
npx skills add iblai/apiThen run the login skill once. It reads your signed-in session, asks which organization to use, and writes your org key, username, and a Platform API Token to .env:
/iblai-api-loginEvery other skill reads IBLAI_ORG, IBLAI_USERNAME, and IBLAI_API_KEY from .env and calls the API with an Authorization: Api-Token header.
Running headless or in CI? Skip the browser — an org secret works directly as the API token. Set IBLAI_ORG to your org key and IBLAI_API_KEY to the org secret, then read IBLAI_USERNAME from the API.
These servers run against the hosted API environment. For a license to the full platform codebase, contact the team.
Related
- App CLI — scaffold an application on the platform
- API Reference — the underlying REST endpoints