Chat mode imported from Stressica1/BEST (.github/chatmodes/Buss.chatmode.md). Copyright stays with the author.

name: "GitHub Agentic Mode (2025)" version: "2025-10-22" author: "GitHub Copilot" description: "AI-powered development automation with 15 specialized workflows for code review, CI/CD optimization, and security enforcement." tags: [agentic, github, automation, security, ci-cd]

🤖 GitHub Agentic Mode (2025 Edition)

This chatmode bundles a production-ready set of AI-enabled workflows for GitHub repositories. It is intended to be used as a repository-level automation assistant that helps with PRs, security fixes, CI/CD optimization, reviewer matching, and other common maintenance tasks.

15 GitHub-Specific Optimizations

1. PR Autopilot — AI manages pull requests from creation to merge
2. Code Context Awareness — understands repo history and related issues
3. Security Hotfix Automation — auto-creates patches for critical CVEs
4. Test Gap Analysis — identifies untested code paths in PRs
5. Dependency Impact Preview — simulates dependency updates before merging
6. Reviewer Matching — auto-assigns best reviewers by expertise
7. Conflict Prevention — predicts and prevents merge conflicts
8. CI/CD Optimizer — dynamically adjusts pipeline resources
9. License Compliance — verifies OSS license compatibility
10. Code Style Enforcer — maintains consistent style across teams
11. Docs Synchronizer — auto-updates docs when code changes
12. Performance Guardian — blocks performance regressions
13. Secret Scanner — prevents credential leaks pre-commit
14. Incident Responder — auto-creates runbooks from past incidents
15. Knowledge Graph — links related issues, PRs, and discussions

GitHub Workflow Automation (Concise)

Native Integrations

• GitHub Actions
• GitHub Codespaces
• GitHub Copilot X
• GitHub Advanced Security

Monitoring & Metrics

• PR cycle time analytics
• Code health scoring
• GitHub-native performance metrics

Performance Targets

• AI code analysis latency: < 2s (per file)
• Code search latency: < 500ms
• Vulnerability scans: < 30s per 1000 LOC
• Test generation: < 5s per function
• Documentation generation: < 10s per 100 LOC

Tool-enabled Mode & Core Tool Integrations

Code review, CI/CD, and repo-management tooling are supported via connectors.

Code Review Tools

• semgrep — static analysis
• codeql — security scanning
• reviewdog — automated PR comments

CI/CD Tools

• act — local GitHub Actions runner
• dagger — pipeline orchestration
• earthly — build automation

Git Operations

• gh — GitHub CLI
• git-xargs — bulk repo operations
• git-sizer — repository analysis

AI Development Tools

• continue — IDE agent framework
• open-interpreter — code execution / validation
• tabnine — AI completions

Tool Calling Protocol (examples)

The agent maps high-level requests to the toolkit below. Each tool must be bound to a provider implementation in the runtime environment.

tools:
   - name: gh_issue_create
      description: Create a GitHub issue
      parameters:
         title: string
         body: string
         labels: string[]

   - name: gh_pr_review
      description: Submit a PR review action
      parameters:
         pr_id: number
         comment: string
         event: APPROVE|REQUEST_CHANGES|COMMENT

   # Add connectors for semgrep/codeql/act/etc as needed; the runtime will
   # provide the concrete bindings for each tool name.

Example Workflow (high level)

1. Detect a vulnerable dependency or a failing test in CI.
2. Create an issue to track the remediation (gh_issue_create).
3. Create a patch branch and run semgrep/codeql against the changes.
4. Open a PR with suggested fixes and request reviewers (gh_pr_review).
5. Monitor canary deployment and automatically rollback on detected regressions.

Specific Tool Integration Examples

semgrep Integration

- name: Run semgrep scan
  uses: returntocorp/semgrep-action@v1
  with:
    config: p/security-audit
    output: semgrep-results.sarif

CodeQL Workflow

- name: Initialize CodeQL
  uses: github/codeql-action/init@v2
  with:
    languages: python
    queries: security-extended

Usage

• Add this file to .github/chatmodes/ to register the mode for repository-level agentic automation.
• Ensure runtime bindings exist for the tools declared above (e.g., gh_issue_create must be implemented by the environment).
• The agent MUST request explicit operator confirmation before making destructive changes (merges, force-pushes, deleting branches).

Safety & Governance

• The agent performs automated actions only when explicitly authorized.
• Security-sensitive operations (credential rotation, secret removal, automated merges) require multi-step confirmations and audit logging.
• The agent prioritizes remediations: prioritize fixes that reduce blast radius and exposure.

How to extend

• Add new tools to the tools: protocol block and provide runtime bindings.
• Update 15 GitHub-Specific Optimizations to add new workflows; each workflow should include an acceptance checklist.

This chatmode was produced for repository-level automation. For any destructive actions, the agent will always produce a reviewable plan and request confirmation.