PR Intent Checker GitHub Action (MCP Server Packaged)

@kevinjcwu

リポジトリ

基本情報

カテゴリ

バージョン管理

トランスポート

stdio

作者

kevinjcwu

PR Intent Checker GitHub Action について

PR Intent Checker action with packaged MCP server

PR Intent Checker GitHub Action の設定

標準の設定はありません

このサーバーの README には解析可能な MCP 設定ブロックが含まれていません。インストール手順はリポジトリをご確認ください。

リポジトリ

PR Intent Checker GitHub Action のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is PR Intent Checker GitHub Action (MCP Server Packaged)?

A GitHub Action that uses Azure OpenAI to analyze a PR’s code changes against requirements in a linked GitHub Issue, detecting intent drift. It packages the GitHub MCP server inside the Docker container to interact with the GitHub API.

How to use PR Intent Checker GitHub Action (MCP Server Packaged)?

Add the action to a workflow with the required permissions and inputs (github_token, azure_openai_endpoint, azure_openai_key, azure_openai_deployment). Configure the corresponding secrets in your repository. Link the relevant issue in the PR description using one of the supported keywords (e.g., Closes #<number>).

Key features of PR Intent Checker GitHub Action (MCP Server Packaged)

Packages GitHub MCP server inside Docker.
Analyzes PR diffs using Azure OpenAI.
Compares code changes against linked issue requirements.
Posts AI results as PR comments.
Outputs PASS/FAIL with explanation.

Use cases of PR Intent Checker GitHub Action (MCP Server Packaged)

Verify PR code matches issue requirements before merge.
Catch early intent drift in code changes.
Automate code review for requirement alignment.
Enforce consistency between PRs and project goals.

よくある質問

How does the action link a PR to an issue?

It scans the PR description for keywords like `Closes`, `Fixes`, or `Resolves` followed by the issue number (e.g., `Closes #4`). It also checks timeline events for linked issues.

What Azure OpenAI credentials are required?

You must provide three secrets: `AZURE_OPENAI_ENDPOINT` (your endpoint URL), `AZURE_OPENAI_KEY` (API key), and `AZURE_OPENAI_DEPLOYMENT` (the model deployment name).

What permissions does the workflow need?

The workflow requires `contents: read`, `pull-requests: write`, and `issues: read`.

What are the supported issue linking formats?

Formats include `Closes #<number>`, `Closes: #<number>`, `Fixes #<number>`, `Fixes: #<number>`, `Resolves #<number>`, `Resolves: #<number>`, and their past-tense variants (e.g., `Closed`, `Fixed`, `Resolved`). Case is ignored.

How does the action communicate with the GitHub API?

The Python script inside the Docker container communicates with the background MCP server via stdin/stdout using JSON-RPC messages. It does not rely on external libraries like PyGithub.