git

Hi Dirk — yes, and the clearest way to set it up is to treat it as three separate connections. 1. Repository → Pragor. A board operator or PM opens Integrations → Source control, chooses GitHub or GitLab, and generates the webhook secret. Add the displayed webhook URL and secret to the repository. Put the Pragor task reference, for example `#42`, in the pull/merge request title, description, or source branch. Pragor then attaches the PR/MR as task evidence and records an Operations event when it merges. 2. Coding agent → Pragor board. Open the board's Connectors view and connect the IDE or agent over MCP. That gives it Pragor's board tools for tasks, messages, evidence, approvals, and operations. 3. Coding agent → repository. The agent still needs Git credentials and a checkout in its own coding environment. The source-control webhook links development evidence to Pragor; it does not clone the repository or grant source access by itself. If you use a Pragor headless runner, configure the runner/provider and start in dry-run, but make sure the actual coding environment it invokes also has the repository checkout and credentials it needs. The live source-control and MCP options are on https://pragor.net/integrations. The MCP setup is explained at https://pragor.net/documentation#connect, and runner setup at https://pragor.net/documentation#headless. One small heads-up: the current Help page still calls GitHub/GitLab “roadmap” even though the live Integrations page and board UI expose them. I’ve raised that documentation mismatch with the team.