adding webhook to check if pr author and commiters are part of approv…

.github/workflows/pr-codebuild-trigger.yml

@@ -0,0 +1,256 @@
+name: PR CodeBuild Trigger (Privileged)
+
+on:
+  workflow_run:
+    workflows: ["PR Team Check (Untrusted)"]
+    types:
+      - completed
+
+permissions:
+  contents: read
+  pull-requests: write
+  issues: write
+
+jobs:
+  trigger-codebuild:
+    runs-on: ubuntu-latest
+    if: >
+      github.event.workflow_run.event == 'pull_request' &&
+      github.event.workflow_run.conclusion == 'success'
+    steps:
+      - name: Download PR information artifact
+        uses: actions/github-script@v7
+        with:
+          script: |
+            var artifacts = await github.rest.actions.listWorkflowRunArtifacts({
+                owner: context.repo.owner,
+                repo: context.repo.repo,
+                run_id: ${{ github.event.workflow_run.id }},
+            });
+            var matchArtifact = artifacts.data.artifacts.filter((artifact) => {
+              return artifact.name == "pr-info";
+            })[0];
+            var download = await github.rest.actions.downloadArtifact({
+                owner: context.repo.owner,
+                repo: context.repo.repo,
+                artifact_id: matchArtifact.id,
+                archive_format: 'zip',
+            });
+            var fs = require('fs');
+            fs.writeFileSync('${{ github.workspace }}/pr-info.zip', Buffer.from(download.data));
+
+      - name: Extract and validate PR information
+        id: extract-pr-info
+        run: |
+          mkdir -p tmp
+          unzip -d tmp/ pr-info.zip
+
+          # Read and validate PR information
+          PR_NUMBER=$(cat ./tmp/pr_number)
+          PR_AUTHOR=$(cat ./tmp/pr_author)
+          COMMIT_AUTHORS=$(cat ./tmp/commit_authors)
+          COMMIT_USERNAMES=$(cat ./tmp/commit_usernames)
+          WORKFLOW_MODIFIED=$(cat ./tmp/workflow_modified)
+          WORKFLOW_CHANGES=$(cat ./tmp/workflow_changes)
+          HEAD_SHA=$(cat ./tmp/head_sha)
+          BASE_SHA=$(cat ./tmp/base_sha)
+          REPOSITORY=$(cat ./tmp/repository)
+
+          # Validate that PR_NUMBER is numeric
+          if ! [[ "$PR_NUMBER" =~ ^[0-9]+$ ]]; then
+            echo "::error::Invalid PR number: $PR_NUMBER"
+            exit 1
+          fi
+
+          # Validate SHA format (40 character hex)
+          if ! [[ "$HEAD_SHA" =~ ^[a-f0-9]{40}$ ]]; then
+            echo "::error::Invalid HEAD SHA format: $HEAD_SHA"
+            exit 1
+          fi
+
+          if ! [[ "$BASE_SHA" =~ ^[a-f0-9]{40}$ ]]; then
+            echo "::error::Invalid BASE SHA format: $BASE_SHA"
+            exit 1
+          fi
+
+          # Validate repository format
+          if ! [[ "$REPOSITORY" =~ ^[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+$ ]]; then
+            echo "::error::Invalid repository format: $REPOSITORY"
+            exit 1
+          fi
+
+          # Set outputs
+          echo "pr_number=$PR_NUMBER" >> $GITHUB_OUTPUT
+          echo "pr_author=$PR_AUTHOR" >> $GITHUB_OUTPUT
+          echo "commit_authors=$COMMIT_AUTHORS" >> $GITHUB_OUTPUT
+          echo "commit_usernames=$COMMIT_USERNAMES" >> $GITHUB_OUTPUT
+          echo "workflow_modified=$WORKFLOW_MODIFIED" >> $GITHUB_OUTPUT
+          echo "workflow_changes=$WORKFLOW_CHANGES" >> $GITHUB_OUTPUT
+          echo "head_sha=$HEAD_SHA" >> $GITHUB_OUTPUT
+          echo "base_sha=$BASE_SHA" >> $GITHUB_OUTPUT
+          echo "repository=$REPOSITORY" >> $GITHUB_OUTPUT
+
+          echo "✅ PR information validated successfully"
+          echo "PR Number: $PR_NUMBER"
+          echo "PR Author: $PR_AUTHOR"
+          echo "Workflow Modified: $WORKFLOW_MODIFIED"
+
+      - name: Security validation
+        id: security-check
+        run: |
+          if [[ "${{ steps.extract-pr-info.outputs.workflow_modified }}" == "true" ]]; then
+            echo "🚨 SECURITY BLOCK: This PR modifies workflow files"
+            echo "Modified files: ${{ steps.extract-pr-info.outputs.workflow_changes }}"
+
+            # Post security block comment on PR
+            curl -X POST \
+              -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
+              -H "Accept: application/vnd.github.v3+json" \
+              https://api.github.com/repos/${{ steps.extract-pr-info.outputs.repository }}/issues/${{ steps.extract-pr-info.outputs.pr_number }}/comments \
+              -d "{\"body\":\"🚨 **SECURITY BLOCK**\\n\\n**CodeBuild execution has been BLOCKED** because this PR modifies GitHub Actions workflow files:\\n\`\`\`\\n${{ steps.extract-pr-info.outputs.workflow_changes }}\\n\`\`\`\\n\\n**Security Policy:** PRs that modify workflows cannot trigger automated builds to prevent security bypass attacks.\\n\\n**Required Actions:**\\n1. Get approval from code owners (@awslabs/sagemaker-1p-algorithms)\\n2. Manual review of all workflow changes\\n3. Separate the workflow changes into a different PR if needed\\n\\n**This is an automated security measure to protect AWS resources.**\"}"
+
+            echo "security_blocked=true" >> $GITHUB_OUTPUT
+          else
+            echo "✅ Security validation passed"
+            echo "security_blocked=false" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Check team membership for PR author
+        id: check-pr-author
+        if: steps.security-check.outputs.security_blocked == 'false'
+        run: |
+          TEAM_MEMBER_FOUND=false
+
+          # Check PR author team membership
+          echo "Checking team membership for PR author: ${{ steps.extract-pr-info.outputs.pr_author }}"
+
+          # Check if PR author is in the team
+          response=$(curl -s -w "%{http_code}" -o /tmp/team_check \
+            -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
+            -H "Accept: application/vnd.github.v3+json" \
+            "https://api.github.com/orgs/awslabs/teams/sagemaker-1p-algorithms/members/${{ steps.extract-pr-info.outputs.pr_author }}")
+
+          if [[ "$response" == "204" ]]; then
+            echo "✅ PR author ${{ steps.extract-pr-info.outputs.pr_author }} is a member of sagemaker-1p-algorithms team"
+            TEAM_MEMBER_FOUND=true
+          elif [[ "$response" == "404" ]]; then
+            echo "❌ PR author ${{ steps.extract-pr-info.outputs.pr_author }} is not a member of sagemaker-1p-algorithms team"
+          else
+            echo "⚠️  Unable to verify team membership for PR author (HTTP $response)"
+            cat /tmp/team_check
+          fi
+
+          echo "pr_author_is_member=$TEAM_MEMBER_FOUND" >> $GITHUB_OUTPUT
+
+      - name: Check team membership for commit authors
+        id: check-commit-authors
+        if: steps.security-check.outputs.security_blocked == 'false'
+        run: |
+          TEAM_MEMBER_FOUND=false
+
+          # Check commit authors if we have usernames
+          if [[ -n "${{ steps.extract-pr-info.outputs.commit_usernames }}" ]]; then
+            IFS=',' read -ra USERNAMES <<< "${{ steps.extract-pr-info.outputs.commit_usernames }}"
+            for username in "${USERNAMES[@]}"; do
+              if [[ -n "$username" ]]; then
+                echo "Checking team membership for commit author: $username"
+
+                response=$(curl -s -w "%{http_code}" -o /tmp/team_check_commit \
+                  -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
+                  -H "Accept: application/vnd.github.v3+json" \
+                  "https://api.github.com/orgs/awslabs/teams/sagemaker-1p-algorithms/members/$username")
+
+                if [[ "$response" == "204" ]]; then
+                  echo "✅ Commit author $username is a member of sagemaker-1p-algorithms team"
+                  TEAM_MEMBER_FOUND=true
+                  break
+                elif [[ "$response" == "404" ]]; then
+                  echo "❌ Commit author $username is not a member of sagemaker-1p-algorithms team"
+                else
+                  echo "⚠️  Unable to verify team membership for commit author $username (HTTP $response)"
+                fi
+              fi
+            done
+          else
+            echo "No GitHub usernames found in commit authors"
+          fi
+
+          echo "commit_author_is_member=$TEAM_MEMBER_FOUND" >> $GITHUB_OUTPUT
+
+      - name: Configure AWS Credentials
+        if: |
+          steps.security-check.outputs.security_blocked == 'false' && 
+          (steps.check-pr-author.outputs.pr_author_is_member == 'true' || steps.check-commit-authors.outputs.commit_author_is_member == 'true')
+        uses: aws-actions/configure-aws-credentials@v4
+        with:
+          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+          aws-region: us-west-2
+
+      - name: Trigger CodeBuild Projects
+        if: |
+          steps.security-check.outputs.security_blocked == 'false' && 
+          (steps.check-pr-author.outputs.pr_author_is_member == 'true' || steps.check-commit-authors.outputs.commit_author_is_member == 'true')
+        run: |
+          echo "🚀 Team member found! Triggering CodeBuild projects..."
+
+          # Trigger tgi-pr-GPU CodeBuild project
+          echo "Starting tgi-pr-GPU build..."
+          TGI_BUILD_ID=$(aws codebuild start-build \
+            --project-name tgi-pr-GPU \
+            --source-version ${{ steps.extract-pr-info.outputs.head_sha }} \
+            --environment-variables-override \
+              name=GITHUB_PR_NUMBER,value=${{ steps.extract-pr-info.outputs.pr_number }} \
+              name=GITHUB_PR_HEAD_SHA,value=${{ steps.extract-pr-info.outputs.head_sha }} \
+              name=GITHUB_PR_BASE_SHA,value=${{ steps.extract-pr-info.outputs.base_sha }} \
+              name=GITHUB_REPOSITORY,value=${{ steps.extract-pr-info.outputs.repository }} \
+            --query 'build.id' --output text)
+
+          echo "TGI CodeBuild started with ID: $TGI_BUILD_ID"
+
+          # Trigger tei-pr-CPU CodeBuild project
+          echo "Starting tei-pr-CPU build..."
+          TEI_BUILD_ID=$(aws codebuild start-build \
+            --project-name tei-pr-CPU \
+            --source-version ${{ steps.extract-pr-info.outputs.head_sha }} \
+            --environment-variables-override \
+              name=GITHUB_PR_NUMBER,value=${{ steps.extract-pr-info.outputs.pr_number }} \
+              name=GITHUB_PR_HEAD_SHA,value=${{ steps.extract-pr-info.outputs.head_sha }} \
+              name=GITHUB_PR_BASE_SHA,value=${{ steps.extract-pr-info.outputs.base_sha }} \
+              name=GITHUB_REPOSITORY,value=${{ steps.extract-pr-info.outputs.repository }} \
+            --query 'build.id' --output text)
+
+          echo "TEI CodeBuild started with ID: $TEI_BUILD_ID"
+
+          # Create a comment on the PR with build information
+          curl -X POST \
+            -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
+            -H "Accept: application/vnd.github.v3+json" \
+            https://api.github.com/repos/${{ steps.extract-pr-info.outputs.repository }}/issues/${{ steps.extract-pr-info.outputs.pr_number }}/comments \
+            -d "{\"body\":\"🚀 **CodeBuild Triggered**\\n\\n✅ Team member verification passed\\n\\n**Build IDs:**\\n- TGI GPU Build: \`$TGI_BUILD_ID\`\\n- TEI CPU Build: \`$TEI_BUILD_ID\`\\n\\nYou can monitor the builds in the [AWS CodeBuild Console](https://us-west-2.console.aws.amazon.com/codesuite/codebuild/projects).\"}"
+
+      - name: Team membership check failed
+        if: |
+          steps.security-check.outputs.security_blocked == 'false' && 
+          steps.check-pr-author.outputs.pr_author_is_member == 'false' && 
+          steps.check-commit-authors.outputs.commit_author_is_member == 'false'
+        run: |
+          echo "❌ Access denied: Neither PR author nor commit authors are members of the sagemaker-1p-algorithms team"
+
+          # Create a comment on the PR about the failed check
+          curl -X POST \
+            -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
+            -H "Accept: application/vnd.github.v3+json" \
+            https://api.github.com/repos/${{ steps.extract-pr-info.outputs.repository }}/issues/${{ steps.extract-pr-info.outputs.pr_number }}/comments \
+            -d "{\"body\":\"❌ **CodeBuild Access Denied**\\n\\nThe PR author and commit authors are not members of the \`sagemaker-1p-algorithms\` team.\\n\\n**Checked:**\\n- PR Author: @${{ steps.extract-pr-info.outputs.pr_author }}\\n- Commit Authors: ${{ steps.extract-pr-info.outputs.commit_authors }}\\n\\nPlease ensure you are a member of the required team to trigger builds.\"}"
+
+          exit 1
+
+      - name: Security block failure
+        if: steps.security-check.outputs.security_blocked == 'true'
+        run: |
+          echo "🚨 SECURITY BLOCK: Workflow execution terminated due to workflow file modifications"
+          echo "::error::SECURITY POLICY VIOLATION: This PR modifies GitHub Actions workflows and has been blocked from executing CodeBuild projects."
+          echo "::error::This is a security measure to prevent malicious workflow modifications from bypassing team membership checks."
+          echo "::error::Please get approval from code owners (@awslabs/sagemaker-1p-algorithms) and consider separating workflow changes into a different PR."
+          exit 1

.github/workflows/pr-team-check.yml

@@ -0,0 +1,85 @@
+name: PR Team Check (Untrusted)
+
+on:
+  pull_request:
+    types: [opened, synchronize, reopened]
+
+permissions:
+  contents: read
+
+jobs:
+  check-team-membership:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout PR code (untrusted)
+        uses: actions/checkout@v4
+        with:
+          ref: ${{ github.event.pull_request.head.sha }}
+          fetch-depth: 0
+
+      - name: Get PR and commit authors
+        id: get-authors
+        run: |
+          # Get PR author
+          PR_AUTHOR="${{ github.event.pull_request.user.login }}"
+          echo "pr_author=$PR_AUTHOR" >> $GITHUB_OUTPUT
+          echo "PR Author: $PR_AUTHOR"
+
+          # Get all commit authors in this PR
+          git fetch origin ${{ github.event.pull_request.base.ref }}
+          COMMIT_AUTHORS=$(git log origin/${{ github.event.pull_request.base.ref }}..${{ github.event.pull_request.head.sha }} --format="%an" | sort -u | tr '\n' ',' | sed 's/,$//')
+          echo "commit_authors=$COMMIT_AUTHORS" >> $GITHUB_OUTPUT
+          echo "Commit Authors: $COMMIT_AUTHORS"
+
+          # Get all commit author usernames (GitHub usernames)
+          COMMIT_USERNAMES=""
+          for sha in $(git log origin/${{ github.event.pull_request.base.ref }}..${{ github.event.pull_request.head.sha }} --format="%H"); do
+            author_email=$(git show --format="%ae" --no-patch $sha)
+            # Try to get GitHub username from commit
+            if [[ "$author_email" == *"@users.noreply.github.com" ]]; then
+              username=$(echo "$author_email" | cut -d'@' -f1 | sed 's/^[0-9]*+//')
+              COMMIT_USERNAMES="$COMMIT_USERNAMES,$username"
+            fi
+          done
+          COMMIT_USERNAMES=$(echo "$COMMIT_USERNAMES" | sed 's/^,//' | sed 's/,$//')
+          echo "commit_usernames=$COMMIT_USERNAMES" >> $GITHUB_OUTPUT
+          echo "Commit Usernames: $COMMIT_USERNAMES"
+
+      - name: Check for workflow modifications
+        id: check-workflow-changes
+        run: |
+          echo "🔒 Security Check: Validating PR does not modify workflow files"
+
+          # Check if PR modifies any workflow files
+          WORKFLOW_CHANGES=$(git diff --name-only origin/${{ github.event.pull_request.base.ref }}...${{ github.event.pull_request.head.sha }} | grep -E '^\.github/workflows/' || true)
+
+          if [[ -n "$WORKFLOW_CHANGES" ]]; then
+            echo "🚨 SECURITY ALERT: This PR modifies workflow files:"
+            echo "$WORKFLOW_CHANGES"
+            echo "workflow_modified=true" >> $GITHUB_OUTPUT
+            echo "workflow_changes=$WORKFLOW_CHANGES" >> $GITHUB_OUTPUT
+          else
+            echo "✅ No workflow files modified in this PR"
+            echo "workflow_modified=false" >> $GITHUB_OUTPUT
+            echo "workflow_changes=" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Save PR information for privileged workflow
+        run: |
+          mkdir -p ./pr-info
+          echo "${{ github.event.pull_request.number }}" > ./pr-info/pr_number
+          echo "${{ steps.get-authors.outputs.pr_author }}" > ./pr-info/pr_author
+          echo "${{ steps.get-authors.outputs.commit_authors }}" > ./pr-info/commit_authors
+          echo "${{ steps.get-authors.outputs.commit_usernames }}" > ./pr-info/commit_usernames
+          echo "${{ steps.check-workflow-changes.outputs.workflow_modified }}" > ./pr-info/workflow_modified
+          echo "${{ steps.check-workflow-changes.outputs.workflow_changes }}" > ./pr-info/workflow_changes
+          echo "${{ github.event.pull_request.head.sha }}" > ./pr-info/head_sha
+          echo "${{ github.event.pull_request.base.sha }}" > ./pr-info/base_sha
+          echo "${{ github.repository }}" > ./pr-info/repository
+
+      - name: Upload PR information artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: pr-info
+          path: pr-info/
+          retention-days: 1

CODEOWNERS

@@ -1 +1,21 @@
-* @awslabs/sagemaker-1p-algorithms
+# Code Owners for llm-hosting-container repository
+
+# Global fallback - require review from sagemaker-1p-algorithms team
+* @awslabs/sagemaker-1p-algorithms
+
+# GitHub Actions workflows - CRITICAL SECURITY
+# These files control CI/CD and AWS resource access
+# Require mandatory review from the sagemaker-1p-algorithms team
+.github/workflows/ @awslabs/sagemaker-1p-algorithms
+.github/actions/ @awslabs/sagemaker-1p-algorithms
+
+# Security-sensitive configuration files
+CODEOWNERS @awslabs/sagemaker-1p-algorithms
+.gitignore @awslabs/sagemaker-1p-algorithms
+
+# Documentation that affects security setup
+docs/github-webhook-setup.md @awslabs/sagemaker-1p-algorithms
+docs/webhook-security-analysis.md @awslabs/sagemaker-1p-algorithms
+
+# AWS CodeBuild specifications
+**/buildspec.yml @awslabs/sagemaker-1p-algorithms