Add PR Summary step

.github/workflows/format.yml

@@ -9,6 +9,7 @@ on:
     branches: [main, master]
   pull_request:
     branches: [main, master]
+    types: [opened, closed, synchronize]
 
 jobs:
   format:
@@ -17,8 +18,12 @@ jobs:
       - name: Run Ultralytics Formatting
         uses: ultralytics/actions@main
         with:
-          token: ${{ secrets.GITHUB_TOKEN }} # automatically generated
+          token: ${{ secrets.GITHUB_TOKEN }} # automatically generated, do not modify
           python: true
-          prettier: true
+          markdown: true
           spelling: false
-          links: true
+          links: false
+          summary: true  # requires either 'openai_api_key' or 'openai_azure_api_key' and 'openai_azure_endpoint'
+          openai_azure_api_key: ${{ secrets.OPENAI_AZURE_API_KEY }}
+          openai_azure_endpoint: ${{ secrets.OPENAI_AZURE_ENDPOINT }}
+          #openai_api_key: ${{ secrets.OPENAI_API_KEY }}

README.md

@@ -50,7 +50,6 @@ To use this action in your Ultralytics repository:
            with:
              token: ${{ secrets.GITHUB_TOKEN }} # automatically generated
              python: true
-             docstrings: true
              markdown: true
              spelling: true
              links: true

action.yml

@@ -28,13 +28,32 @@ inputs:
     description: "Run Broken Links checks"
     required: false
     default: "false"
+  summary:
+    description: "Run PR Summary"
+    required: false
+    default: "false"
+  openai_api_key:
+    description: "OpenAI API Key"
+    required: false
+  openai_azure_api_key:
+    description: "OpenAI Azure API Key"
+    required: false
+  openai_azure_endpoint:
+    description: "OpenAI Azure Endpoint"
+    required: false
+  openai_model:
+    description: "OpenAI Model"
+    required: false
+    default: "gpt-4-1106-preview"
 runs:
   using: "composite"
   steps:
     - name: Print Action Information
       run: |
-        echo "github.event_name: ${{ github.event_name }}"
+        echo "github.event_name: ${{ github.event_name }}"  # i.e. "pull_request"
+        echo "github.event.action: ${{ github.event.action }}"  # i.e. "open" or "close"
         echo "github.repository: ${{ github.repository }}"
+        echo "github.event.pull_request.number: ${{ github.event.pull_request.number }}"
         echo "github.event.pull_request.head.repo.full_name: ${{ github.event.pull_request.head.repo.full_name }}"
         echo "github.actor: ${{ github.actor }}"
         echo "github.event.pull_request.head.ref: ${{ github.event.pull_request.head.ref }}"
@@ -88,7 +107,7 @@ runs:
 
     # Prettier (JavaScript, JSX, Angular, Vue, Flow, TypeScript, CSS, HTML, JSON, GraphQL, Markdown, YAML) -------------
     - name: Run Prettier
-      if: inputs.prettier == 'true' || inputs.markdown == 'true'
+      if: inputs.prettier == 'true'
       run: |
         npm install --global prettier
         npx prettier \
@@ -123,6 +142,23 @@ runs:
       shell: bash
       continue-on-error: false
 
+    # PR Summary -------------------------------------------------------------------------------------------------------
+    - name: Summarize PR
+      if: inputs.summary == 'true' && (github.event_name == 'pull_request' || github.event_name == 'pull_request_target') && (inputs.event.action == 'opened' || inputs.event.action == 'closed')
+      env:
+        REPO_NAME: ${{ github.repository }}
+        PR_NUMBER: ${{ github.event.pull_request.number }}
+        GITHUB_TOKEN: ${{ inputs.token }}
+        OPENAI_API_KEY: ${{ inputs.openai_api_key }}
+        OPENAI_AZURE_API_KEY: ${{ inputs.openai_azure_api_key }}
+        OPENAI_AZURE_ENDPOINT: ${{ inputs.openai_azure_endpoint }}
+        OPENAI_MODEL: ${{ inputs.openai_model }}
+      run: |
+        pip install --no-cache -q openai
+        python utils/run_pr_summary.py
+      shell: bash
+      continue-on-error: true
+
     # Broken links -----------------------------------------------------------------------------------------------------
     - name: Broken Link Checker
       if: inputs.links == 'true'

utils/run_pr_summary.py

@@ -0,0 +1,107 @@
+import os
+
+import requests
+from openai import OpenAI, AzureOpenAI
+
+REPO_NAME = os.getenv("REPO_NAME")
+PR_NUMBER = os.getenv("PR_NUMBER")
+GITHUB_TOKEN = os.getenv("GITHUB_TOKEN")
+GITHUB_HEADERS = {"Authorization": f"token {GITHUB_TOKEN}"}
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+OPENAI_AZURE_API_KEY = os.getenv("OPENAI_AZURE_API_KEY")
+OPENAI_AZURE_ENDPOINT = os.getenv("OPENAI_AZURE_ENDPOINT")
+OPENAI_AZURE_BOTH = OPENAI_AZURE_API_KEY and OPENAI_AZURE_ENDPOINT
+OPENAI_MODEL = os.getenv("OPENAI_MODEL")
+OPENAI_MODEL_TOKENS = 128000  # update with model
+SUMMARY_START = (
+    "## 🛠️ PR Summary\n\n<sub>Made with ❤️ by [Ultralytics Actions](https://github.com/ultralytics/actions)<sub>\n\n"
+)
+
+# Checks
+assert OPENAI_MODEL, "No model found, please define OPENAI_MODEL"
+assert (
+    OPENAI_API_KEY or OPENAI_AZURE_BOTH
+), "No OpenAI Keys found, please pass either OPENAI_API_KEY or both (OPENAI_AZURE_API_KEY and OPENAI_AZURE_ENDPOINT)"
+if OPENAI_AZURE_API_KEY or OPENAI_AZURE_ENDPOINT:
+    assert OPENAI_AZURE_BOTH, "For Azure usage both both OPENAI_AZURE_API_KEY and OPENAI_AZURE_ENDPOINT must be passed."
+
+
+def openai_client(azure=OPENAI_AZURE_BOTH):
+    """Returns OpenAI client instance."""
+    return (
+        AzureOpenAI(
+            api_key=OPENAI_AZURE_API_KEY, api_version="2023-09-01-preview", azure_endpoint=OPENAI_AZURE_ENDPOINT
+        )
+        if azure
+        else OpenAI(api_key=OPENAI_API_KEY)
+    )
+
+
+def get_pr_diff(repo_name, pr_number):
+    """Fetches the diff of a specific PR from a GitHub repository."""
+    url = f"https://github.com/gitapi/repos/{repo_name}/pulls/{pr_number}"
+    headers = {"Authorization": f"token {GITHUB_TOKEN}", "Accept": "application/vnd.github.v3.diff"}
+    response = requests.get(url, headers=headers)
+    return response.text if response.status_code == 200 else ""
+
+
+def generate_pr_summary(repo_name, diff_text):
+    """Generates a professionally written yet accessible summary of a PR using OpenAI's API."""
+    if not diff_text:
+        diff_text = "**ERROR: DIFF IS EMPTY, THERE ARE ZERO CODE CHANGES IN THIS PR."
+    ratio = 3.3  # about 3.3 characters per token
+    limit = round(OPENAI_MODEL_TOKENS * ratio * 0.7)  # use up to 70% of the context window
+    messages = [
+        {
+            "role": "system",
+            "content": "You are an Ultralytics AI assistant skilled in software development and technical communication. Your task is to summarize GitHub PRs from Ultralytics in a way that is accurate, concise, and understandable to both expert developers and non-expert users. Focus on highlighting the key changes and their impact in simple, concise terms.",
+        },
+        {
+            "role": "user",
+            "content": f"Summarize this '{repo_name}' PR, focusing on major changes, their purpose, and potential impact. Keep the summary clear and concise, suitable for a broad audience. Add emojis to enliven the summary. Reply directly with a summary along these example guidelines, though feel free to adjust as appropriate:\n\n"
+            f"### 🌟 Summary (single-line synopsis)\n"
+            f"### 📊 Key Changes (bullet points highlighting any major changes)\n"
+            f"### 🎯 Purpose & Impact (bullet points explaining any benefits and potential impact to users)\n"
+            f"\n\nHere's the PR diff:\n\n{diff_text[:limit]}",
+        },
+    ]
+    response = openai_client().chat.completions.create(model=OPENAI_MODEL, messages=messages).choices[0]
+    reply = response.message.content.strip()
+    if len(diff_text) > limit:
+        return SUMMARY_START + "**WARNING ⚠️** this PR is very large, summary may not cover all changes.\n\n" + reply
+    else:
+        return SUMMARY_START + reply
+
+
+def update_pr_description(repo_name, pr_number, new_summary):
+    """Updates the original PR description with a new summary, replacing an existing summary if found."""
+    # Fetch the current PR description
+    pr_url = f"https://github.com/gitapi/repos/{repo_name}/pulls/{pr_number}"
+    pr_response = requests.get(pr_url, headers=GITHUB_HEADERS)
+    pr_data = pr_response.json()
+    current_description = pr_data.get("body") or ""  # warning, can be None
+
+    # Check if existing summary is present and update accordingly
+    if SUMMARY_START in current_description:
+        updated_description = current_description.split(SUMMARY_START)[0] + new_summary
+    else:
+        updated_description = current_description + "\n\n" + new_summary
+
+    # Update the PR description
+    update_response = requests.patch(pr_url, json={"body": updated_description}, headers=GITHUB_HEADERS)
+    return update_response.status_code
+
+
+if __name__ == "__main__":
+    # Fetch PR details
+    diff = get_pr_diff(REPO_NAME, PR_NUMBER)
+
+    # Generate PR summary
+    summary = generate_pr_summary(REPO_NAME, diff)
+
+    # Update PR description
+    status_code = update_pr_description(REPO_NAME, PR_NUMBER, summary)
+    if status_code == 200:
+        print("PR description updated successfully.")
+    else:
+        print(f"Failed to update PR description. Status code: {status_code}")