Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
41 lines (29 loc) · 1.72 KB

CONTRIBUTING.md

File metadata and controls

41 lines (29 loc) · 1.72 KB
Raw

Contributing to AppScope

Thanks for helping build app intelligence that's honest about what it can't know. First-timers welcome — this repo is deliberately friendly to your first PR.

Your first PR in 10 minutes

git clone https://github.com/Ahad690/open-app-intel && cd open-app-intel
python -m venv .venv && . .venv/bin/activate    # Windows: .venv\Scripts\activate
pip install -r requirements.txt pytest
pytest -q                                        # 73 tests — should be green

Pick a good first issue, comment to claim it, open a PR.

The rules (non-negotiable)

  • Estimates are ranges capped at MEDIUM; HIGH is observed facts only.
  • Ads emit proxies, never USD. Free-app revenue is never invented.
  • Federation shares public anchors only (assert_public_only).

See AGENTS.md for the full contract. A PR that breaks any of these doesn't merge.

What makes a good PR here

  • Small and scoped. One issue, one PR, with a test. Keep pytest -q green.
  • Honest. New public data collectors (with provenance), new deterministic model refinements (that stay ≤ MEDIUM), docs, and federation-guard tests are welcome. "Estimate dollar ad spend" or "make installs HIGH-confidence" is not.

Good areas to contribute

  • New top-chart collectors for additional countries/categories.
  • New review-source adapters (observed counts only).
  • Calibration improvements to the rank→download power law (documented, banded).
  • Federation tooling and anchor-guard test coverage.
  • Docs, examples, and MCP client setup guides.

By contributing you agree code is MIT-licensed and data/docs are CC-BY-4.0, like the repo.