Skip to main content
Gentic exposes a Model Context Protocol (MCP) server that makes your entire project and issue workflow available to any MCP-compatible AI assistant. Instead of switching between your AI tool and the Gentic dashboard, you can ask your assistant to create a bug report, update a project, or queue an agent run — and it will call Gentic’s API on your behalf in real time.

​
MCP endpoint

Point your MCP client at the following URL:
This single endpoint supports every Gentic tool, covering account inspection, project management, the full issue lifecycle, blocking relations between issues, and account-scoped Label organization.
The MCP server uses OAuth 2.0 for authentication. The first time your client connects, it will redirect you to app.gentic.chat so you can authorize access with your Gentic account. No API keys or manual token setup is required.

​
Connect Claude Desktop

Claude Desktop reads MCP server configuration from a JSON file. Add the gentic entry to your claude_desktop_config.json:
claude_desktop_config.json
After saving the file and restarting Claude Desktop, Claude will prompt you to authorize the connection through your Gentic account. Once authorized, all Gentic tools are available in your conversation.

​
Managed coding agents connect automatically

Coding agents that Gentic runs for you — Claude Code and Codex sessions started by a host — get this same MCP server without any setup. Every session, whether it is a fresh run or a follow-up that resumes an earlier one, is started with the Gentic MCP server already configured and authenticated with that host’s credential. That connection resolves to your Gentic account, so the agent working on an issue sees exactly the projects, issues, and labels you do. Banning or deleting a host revokes its MCP access along with its API access, immediately and through the same control. Any MCP servers you configure yourself — in .mcp.json, ~/.codex/config.toml, or your provider’s own settings — are left in place; the Gentic server is added alongside them.

​
Connect other MCP clients

Any MCP-compatible client that accepts a server URL — including Cursor, Continue, and custom MCP hosts — can connect using the same endpoint. Consult your client’s documentation for where to add a new MCP server entry, then supply https://app.gentic.chat/mcp as the URL.

​
What you can do after connecting

Once connected, your AI assistant can:

Manage projects

List, create, update, and delete Gentic projects linked to your GitHub repositories.

Create and update issues

Create bug reports, feature requests, and ideas with titles and agent instructions.

Control issue workflow

Move issues through every workflow status, from draft all the way to completed.

Trigger agent runs

Transition an issue from draft to todo to queue a background AI coding agent run immediately.

Organize with Labels

Create account-scoped Labels, assign them across projects, and filter issues by Label or find unlabeled work.

Sequence work with relations

Declare that one issue blocks another so the host runs a chain of issues in the right order.

​
Example: create a bug report and queue an agent run

You do not need to open the Gentic dashboard at all. Ask your assistant something like:
“Create a bug issue for the login project titled ‘Fix null pointer in auth’ with the body ‘The login endpoint throws NPE when email is missing — add input validation.’ Then move it to todo.”
Your assistant will call create_issue to create the issue in draft state, then call update_issue_status to move it to todo, which queues a background agent run automatically.
Start issues in draft status while you refine the title and body. Move to todo only when you are ready for the agent to begin work — that transition is what triggers the run.

​
Available tools

See the Tools reference for a complete list of the MCP tools, their parameters, and example usage.