Connect the assistant
Copy the connection URL from Settings → Integrations & MCP → MCP Server and add it as a connector in Claude or ChatGPT. The first request opens your instance in the browser: sign in, pick the repositories, tick the permissions.
Hand a change to ProPR from the Claude or ChatGPT conversation you are already in. The assistant files the task through your instance's MCP server, your agents build it on your server, and the pull request discussion comes back into the chat.
Copy the connection URL from Settings → Integrations & MCP → MCP Server and add it as a connector in Claude or ChatGPT. The first request opens your instance in the browser: sign in, pick the repositories, tick the permissions.
Describe the change. The assistant opens a GitHub issue and starts implementation with the model you name. Ask for Ultrafix and auto-merge when you want them.
Add retry with backoff to the webhook sender in acme/payments. Use Claude Opus 5.5 and run Ultrafix to 8/10.
An administrator decides what chat clients may do, down to the repository.
For a larger change, ask for a plan. Refine it in the chat, publish it as GitHub issues, and start the issues you pick with the models you choose.
Plan rate limiting for orders-api. Show me the tasks before you publish anything, then start the first two with GPT-6 Astra.
Ask how it is going and the assistant reports each stage, from issue to finished PR, with running tasks, reviews and checks in one overview.
Close the chat and the work keeps running. A new conversation picks it up from the PR number.
The PR's comments come into the chat, ProPR reviews and their F# findings included. Pick the findings worth fixing and the assistant applies exactly those as a follow-up commit.
Summarize the review on acme/payments#57 and apply F2. Leave the suggestions.
Logs → MCP Log records each request: the tool, the repository, the outcome and the timing. Keys and tokens belong in browser sign-in, never in the conversation. The chat assistant guide covers client setup.
Tasks and plans started from a conversation, running on your server, landing as reviewed PRs in GitHub, with their progress one question away.