Add issue action

Signed-off-by: Shuyang Xin <[email protected]>

.github/workflows/issue_answer.yml

@@ -0,0 +1,38 @@
+name: LangChain Issue Processor
+
+on:
+  issues:
+    types: [opened, reopened]
+
+jobs:
+  process-issue:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v2
+      - name: Set up Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: '3.x'
+      - name: Install dependencies
+        run: |
+          pip install langchain openai
+      - name: Run process.py
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        run: |
+          SCRIPT_OUTPUT=$(python ai/issue_answer.py --github-token $GITHUB_TOKEN --openai-token $OPENAI_API_KEY)
+          echo "$SCRIPT_OUTPUT"
+          echo "SCRIPT_OUTPUT=$SCRIPT_OUTPUT" >> $GITHUB_ENV
+          echo "${{env.SCRIPT_OUTPUT}}"
+      - name: comment result
+        uses: actions/github-script@v5        
+        with:
+          script: |
+            github.rest.issues.createComment({
+              issue_number: context.issue.number,
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              body: `${{env.SCRIPT_OUTPUT}}`,
+            });

.github/workflows/issue_label.yml

@@ -0,0 +1,39 @@
+name: LangChain Label Processor
+
+on:
+  issues:
+    types: [opened, reopened]
+
+jobs:
+  process-issue:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v2
+      - name: Set up Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: '3.x'
+      - name: Install dependencies
+        run: |
+          pip install langchain openai
+      - name: Run process.py
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        run: |
+          LABEL_OUTPUT=bug,document
+          echo "$LABEL_OUTPUT"
+          echo "LABEL_OUTPUT=$LABEL_OUTPUT" >> $GITHUB_ENV
+          echo "${{env.LABEL_OUTPUT}}"
+      - name: comment result
+        uses: actions/github-script@v5
+        with:
+          script: |
+            const labelsToAdd = ['${{ env.LABEL_OUTPUT }}'.split(',')];
+            github.rest.issues.addLabels({
+              issue_number: context.issue.number,
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              labels: labelsToAdd,
+            });

ai/issue_answer.py

@@ -0,0 +1,50 @@
+from langchain.document_loaders.github import GitHubIssuesLoader
+from langchain.chat_models import ChatOpenAI
+from langchain.prompts import ChatPromptTemplate
+from langchain.schema.output_parser import StrOutputParser
+from langchain.schema.runnable import RunnableParallel, RunnablePassthrough
+
+import argparse
+
+def parse_arguments():
+    parser = argparse.ArgumentParser(description="Your program description.")
+    parser.add_argument("--github-token", required=True, help="GitHub access token")
+    parser.add_argument("--openai-token", required=True, help="OpenAI API key")
+    return parser.parse_args()
+
+# Parse command line arguments
+args = parse_arguments()
+
+# Create an instance of the GitHubIssuesLoader class
+loader = GitHubIssuesLoader(
+    repo="XinShuYang/antrea",
+    owner="XinShuYang",
+    access_token=args.github_token,
+    include_prs=False,
+)
+
+# Load the issues of the repository
+issues = loader.load()
+
+issue_data = []
+# Print the title and body of each issue
+for issue in issues:
+    if issue.metadata["number"] == 9 :
+        issue_data.append({'title': issue.metadata['title'], 'body': issue.page_content})
+
+template = """Answer the question based only on the following context:
+    {context}
+
+    """
+prompt = ChatPromptTemplate.from_template(template)
+model = ChatOpenAI(model_name="gpt-3.5-turbo", openai_api_key=args.openai_token)
+output_parser = StrOutputParser()
+setup_and_retrieval = RunnableParallel(
+    {"context": RunnablePassthrough()}
+)
+
+chain = setup_and_retrieval | prompt | model | output_parser
+
+result=chain.invoke(issue_data)
+
+print(result)