GitHub DeepBlame Server

Example Use Cases

• 1

  Find every pull request that modified a specific file (with commit-based pagination and time filters).

• 2

  Compile comprehensive contextual data about PRs — including comments, review comments, reviews, file changes, and linked issues — to inform code explanation or root-cause analysis.

• 3

  Iteratively fetch large sets of PR details using pseudo-pagination to avoid rate/response-size limits while aggregating historical change data for a file.

Description

This MCP server provides tools to go beyond git blame by collecting comprehensive pull request data for a given file, including PR metadata, comments, review comments, reviews, file diffs, and linked issues. It uses the GitHub REST API and requires a GitHub Personal Access Token to operate. The workflow is two-step: list PRs that touched a file (with pagination) and then fetch detailed data for those PRs (with pseudo-pagination). This helps agents assemble rich contextual history around code changes for deeper analysis and troubleshooting.