moss-contribution-radar

AI-powered contribution radar for discovering, scoring, and operationalizing open-source PR opportunities for Mossland builders.

Why this exists

Open-source contributors waste time hunting for issues that are:

• actually actionable,
• aligned with their stack,
• active enough to get reviewed,
• and valuable enough to strengthen their portfolio.

moss-contribution-radar turns that messy search process into a repeatable pipeline:

1. ingest candidate repositories, issues, and PR-friendly signals,
2. score contribution opportunities,
3. generate contributor-facing briefs,
4. track outreach and execution status.

Problem statement

Most contributors discover opportunities through ad-hoc browsing on GitHub, social feeds, and Discord communities. That approach creates three problems:

• Low signal density — too many stale or beginner-trap issues.
• Poor personal fit — issues rarely match the contributor's skills or goals.
• Weak execution flow — even when a good issue is found, there is no clean handoff into action.

Vision

Build a lightweight contribution intelligence platform that helps Mossland builders consistently find and close meaningful OSS contributions.

The platform should answer:

• Which repos are most worth watching this week?
• Which issues are realistically mergeable in the next 7–14 days?
• Which opportunities best match a contributor's skills?
• What is the fastest path from discovery to submitted PR?

MVP

The MVP focuses on a narrow but useful loop:

• Pull repositories and issues from GitHub
• Normalize repository health/activity signals
• Score issues for contribution potential
• Surface a ranked shortlist
• Generate a contribution brief for each shortlisted issue

Core MVP features

• Repository watchlist ingestion
  • manually curated seed repositories
  • topic-based GitHub search support
• Opportunity scoring
  • issue freshness
  • maintainer activity
  • label quality (good first issue, help wanted, etc.)
  • repository responsiveness
  • language/tag match
• Contributor matching
  • lightweight user skill profile
  • preferred languages/domains
  • difficulty and urgency filters
• Contribution brief generation
  • repo context summary
  • issue summary
  • probable implementation area
  • first-step checklist
• Workflow status tracking
  • discovered
  • shortlisted
  • claimed
  • in progress
  • PR opened
  • merged / closed

Proposed architecture

GitHub APIs / curated repo list
            ↓
      Ingestion pipeline
            ↓
   Normalized opportunity store
            ↓
      Scoring + ranking engine
            ↓
  Brief generator / action dashboard

Suggested stack

• Backend: Python + FastAPI
• Data jobs: Python cron / scheduled tasks
• Storage: SQLite for MVP, Postgres later
• Frontend: Next.js or simple React dashboard
• Integrations: GitHub REST/GraphQL APIs
• AI layer: LLM-generated contribution briefs and summaries

Data model overview

repositories

• id
• full_name
• description
• language
• stars
• forks
• open_issues_count
• last_push_at
• health_score

opportunities

• id
• repo_id
• issue_number
• title
• labels
• created_at
• updated_at
• score
• difficulty
• status
• brief_markdown

contributors

• id
• name
• skills
• preferred_languages
• interests
• difficulty_band

Scoring dimensions

A first-pass score can combine:

• repository activity recency
• issue freshness
• comment velocity
• maintainers replying in the last N days
• issue label quality
• issue size / ambiguity heuristics
• contributor-skill overlap

Example formula:

opportunity_score =
  0.25 * repo_health +
  0.20 * maintainer_responsiveness +
  0.20 * issue_actionability +
  0.20 * contributor_fit +
  0.15 * merge_likelihood

Example use cases

1. New contributor onboarding

A beginner wants 3 Python issues that are active, well-scoped, and likely to be reviewed quickly.

2. Portfolio-driven contributor

An experienced builder wants higher-signal repos in AI infra, agents, or data tooling.

3. Community ops

A community lead wants to publish a weekly "best issues to tackle" digest for Mossland Open Devs.

Milestones

Phase 1 — Research and ingestion

• define target personas
• create seed repo watchlist
• implement GitHub ingestion
• persist normalized repo/issue data

Phase 2 — Scoring and ranking

• build heuristics engine
• rank opportunities
• validate results on a small curated set

Phase 3 — Brief generation and workflow

• generate issue briefs
• add claim / status workflow
• export weekly shortlist

Phase 4 — Productization

• simple dashboard
• contributor profiles
• recurring automated updates

Repo structure target

moss-contribution-radar/
├── README.md
├── docs/
│   ├── architecture.md
│   └── plans/
├── app/
│   ├── api/
│   ├── scoring/
│   ├── ingestion/
│   └── models/
├── scripts/
├── tests/
└── data/

Immediate next steps

1. Finalize MVP scope and target persona
2. Write architecture and implementation plan docs
3. Scaffold backend project structure
4. Implement GitHub ingestion for a curated repo list
5. Build first heuristic scoring pass
6. Expose ranked opportunities through a simple API

Repository goals

This repository should become:

• a practical contributor discovery engine,
• a reusable dataset pipeline for OSS opportunity analysis,
• and a community-facing tool for Mossland Open Devs.

License

TBD