Add repository audit workflow for identifying agentic workflow opportunities

[Copilot]

Creates an on-demand workflow that audits repositories to identify automation opportunities using agentic workflows. Designed for FStar but works with any GitHub repository.

Implementation

Workflow structure:

• Manual trigger with configurable repository input (default: FStarLang/FStar)
• 45-minute timeout for comprehensive analysis across 6 phases
• Cache memory persists audit results for trend analysis
• Outputs detailed GitHub Discussion with prioritized recommendations

Analysis phases:

1. Repository metadata and contributor analysis via GitHub API
2. Source code structure, patterns, and technical debt scanning
3. GitHub Actions workflow health and efficiency assessment
4. Issue history pattern recognition and categorization
5. Opportunity identification (daily improvers, event-driven automation)
6. Implementation roadmap with effort/impact prioritization

Key features:

• Clones target repo for deep inspection (structure, docs, build files, code metrics)
• Generates 5-8 prioritized opportunities with implementation sketches for top 3
• Repository-specific recommendations (not generic advice)
• Creates single Discussion per run (closes older audits)

Usage

# Audit default repository (FStar)
gh workflow run repo-audit-analyzer.yml

# Audit any repository
gh workflow run repo-audit-analyzer.yml --field repository=owner/repo

Files

• .github/workflows/repo-audit-analyzer.md - Workflow definition (951 bytes)
• .github/agentics/repo-audit-analyzer.md - Agent prompt, editable without recompilation (22 KB)
• .github/workflows/repo-audit-analyzer.lock.yml - Compiled workflow (54 KB)

Output Structure

Discussion includes:

• Repository overview (language, contributors, structure)
• GitHub Actions health metrics and inefficiencies
• Issue pattern analysis with automation potential
• Prioritized workflow opportunities (High/Medium/Future)
• Implementation roadmap (Quick wins → High impact → Long-term)
• Code quality insights and technical debt indicators

Original prompt

Create an agentic workflow that audits an existing repository and it action workflows and identify potential productivity / improvements using agentic workflows.

• research the project
• do a deep survey of existing github actions
• identify source code patterns
• inspect issue history

Apply patterns to discover daily improvers.

Keep a cache memory to store results as the search will be done repreatidley

Apply analyzis to https://github.com/FStarLang/FStar

Custom agent used: agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[pelikhan]

@copilot ensure the project related handler are handled together, the GitHub_token related handlers, the then the assign to agent