Today we are shipping Connic MCP: a direct connection between your coding agent and your Connic project. Any client that speaks MCP, whether that is Codex, Claude Code, Cursor, or something else, can now read live runs, traces, and logs, ship deployments, and manage your project in the same chat where it writes your agents. The detour through the dashboard is over.
Coding agents already write Connic agents well: the YAML, the tools, the middleware. Operating them was a different story. To find out why a run failed or what a deployment's tests said, you left the editor for the dashboard, read the answer, and typed a summary of it back into the chat. The client that wrote the code could not see the system it was running on. Now it can ask directly. This became practical with the 2026-07-28 MCP release, the first protocol revision we would trust between a coding agent and production. Connic MCP is built on it, with a fallback for clients on the older protocol.
The whole development loop in one chat
Building an agent was already fast on Connic. The slow part was everything around it: checking whether the deployment's tests passed, digging into last night's failed runs, confirming the new version actually behaves better than the old one. Connic MCP moves that work into the conversation you are already having.
An agent that can see real traces writes better fixes than one working from your paraphrase of an error, and one that can watch deployment tests can iterate until they pass instead of handing the checking back to you. Every step it can verify itself is a step you no longer sit in the middle of.
What your agent can do
The connection covers the Connic platform end to end, reading and acting on the same things you see in the dashboard:
| Area | See | Do |
|---|---|---|
| Agents and runs | Runs, traces, logs, input and output previews | Cancel or rerun single and bulk runs |
| Deployments | Deployments, tests, deployment runs | Create, cancel, activate, redeploy |
| Connectors | Definitions, links, runs, stats, logs | Create, update, delete, link agents |
| Retrieval | Namespaces, entries, sources, stats | Upload content, manage and sync sources |
| Database | Stats, collections, inferred schemas, full document queries and counts | Update existing rows; delete documents and collections |
| Sessions | Persistent session state, agent and user identity, timestamps | Search and filter; delete individual or matching sessions |
| Approvals and channels | Pending approvals, routing, channels | Approve or reject calls, manage routing and channels |
| Judges and A/B tests | Judge stats and runs, variants, comparisons | Manage judges and tests, queue evaluations |
| Budgets and audit | Cost summaries, rankings, audit events | Manage budget alerts |
| Environments and governance | Environments, health, AI governance records | Manage environments and governance records |
Connect a client in one command
Setup is one command, then a sign-in with your Connic account in the browser:
# Codex
codex mcp add connic --url https://mcp.connic.co/mcp
codex mcp login connic
# Claude Code
claude mcp add --transport http --scope user connic https://mcp.connic.co/mcp
# Cursor Agent, after installing the Connic plugin
cursor-agent mcp login connicThe recommended setup is the Connic plugin, which bundles this connection with the Connic skill, so your client both knows how to write Connic agents and can operate the project they run in. Follow the AI agent setup guide for the plugin install in Codex, Claude Code, Cursor, and OpenCode, and for connecting any other MCP-compatible client manually.
Install the Connic plugin or connect Connic MCP directly, and let your client work on the system it builds for.
Set up your coding clientYou stay in control
Connecting a client does not hand it your project. When you approve the connection, you choose which project it sees, which environments, and whether it can only read or also act. A client can never do more than your own account can, and if your access changes, its access changes with it. Every connection is listed in project settings with what it is allowed to do and when it was last used, every action lands in the audit log, and revoking takes one click.
Some things are simply not there to misuse. A connected client cannot spend your money, read your credentials, delete your project, or start agent runs. To trigger agents from external systems, use a connector, the production path built for it. There are also no API keys involved anywhere: never paste a Connic API key into an MCP configuration or a chat.
The third MCP surface
Connic now speaks MCP in three places, and they point in different directions:
mcp_servers in its YAML. Tools flow into your agent.Connic MCP does not add tools to a deployed agent and does not expose an agent as a tool. If that is what you need, read the MCP servers guide for agents or see the MCP Server connector reference.
How do I start?
- 1.Install the Connic plugin for your client, or add
https://mcp.connic.co/mcpdirectly. Open the AI agent setup guide for the exact commands. - 2.Approve the connection in the browser: pick the project, the environments, and what the client is allowed to do.
- 3.Ask your agent something you would have opened the dashboard for: the last failed run, the state of a deployment, yesterday's cost by agent.