Add optional data fetching to reduce API calls in N+1 query pattern

[Copilot]

The extraction loop makes 3 additional API calls (commits, reviewers, comments) per PR, creating an N+1 query pattern. For 250 PRs: 250 + (250×3) = 1000 API calls.

Changes

• New environment variables to control data fetching (defaults maintain current behavior):

  • FETCH_COMMITS - fetch commit data per PR
  • FETCH_REVIEWERS - fetch reviewer data per PR
  • FETCH_COMMENTS - fetch comment data per PR

• Conditional fetching in extract_pull_requests():

  • Skip API calls when disabled
  • Provide empty data structures ([] or {"users": [], "teams": []}) to maintain schema compatibility

• Performance impact: Disabling all extra fetching reduces calls from 1000 to 250 (75% reduction) for 250 PRs

Example

# Fetch only PR metadata, skip extras
FETCH_COMMITS=false FETCH_REVIEWERS=false FETCH_COMMENTS=false python3 main.py

# Fetch PRs and reviewers only
FETCH_COMMITS=false FETCH_COMMENTS=false python3 main.py

Other fixes

• Corrected extract_reviewers() return type annotation (dict not list[dict])
• Fixed GITHUB_REPOS env var handling to check for None before calling .split()

---

💡 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.

[dklawren]

We need to get the data every time the script is ran so no need for parameters to turn on or off each API call.