Contribution.dev

Contribution.dev turns the dark output of AI-assisted software work into measurable contribution value. Start with a private agentic-readiness audit, then connect AI spend to engineering outcomes and business outcomes.

Open sourceLocal-firstContribution accountingNo automatic uploads
report.md

contribution analyze --since 90d

Your repo is ready in the small, risky in the large.

Agents have clear paths for focused backend fixes. Larger behavior-changing work needs better validation guidance before review.

B

readiness

Medium

confidence

0

raw uploads

Next action

Document validation and add one adjacent test before the next behavior-changing PR.

Readiness evidence

Repo grade, not person score

#142

Auth redaction tests

Strong

Confidence: High

#137

Profile export shape

Mixed

Confidence: Medium

Phase 1: Agentic readiness

Find where agents will succeed or thrash.

The CLI reads instructions, architecture cues, validation setup, tests, churn, risky paths, and optional workflow notes. It turns local repo evidence into the highest-ROI fixes before the next PR.

Agentic readiness audit

Scan instructions, architecture cues, validation setup, tests, churn, and risky paths locally.

GitHub metadata when requested

Add PRs, reviews, checks, comments, and durability evidence only when you choose.

Durability evidence

See which changes stayed stable, churned, reverted, or lacked enough evidence to trust.

Weakness map

Find patterns around instructions, tests, context bloat, churn, risky paths, review burden, and docs.

Private report

Write report.md with readiness evidence, confidence, limits, and highest-ROI next actions.

Public-safe exports

Create profile and share-card JSON without private paths, raw diffs, spend data, or weaknesses by default.

Start with repo readiness.

Run the CLI locally. Read the readiness evidence. Fix the highest-leverage gap before an agent or reviewer pays for it. Share only the proof you choose.

01

Run the CLI

Scan a local repo or public Git URL without uploading raw code by default.

02

Read readiness evidence

See instruction quality, validation setup, tests, churn, risky paths, and confidence notes.

03

Fix the top gap

Improve the highest-ROI issue before an agent or reviewer spends time on it.

04

Share selectively

Export positive proof only when it is safe, accurate, and worth sharing.

Roadmap

The roadmap is the full ROI loop.

Phase 1 measures repo readiness. Phase 2 connects AI spend to durable engineering outcomes. Phase 3 connects engineering work to business outcomes. The wedge is software because the receipts already exist.

Phase 1

Agentic readiness

Grade repo readiness and show the five highest-ROI improvements for AI-assisted development.

Phase 2

Engineering ROI

Connect token spend, model cost, review attention, churn, reverts, and production survival to durable engineering value.

Phase 3

Product ROI

Connect durable engineering work to adoption, retention, revenue, support, and customer outcomes.

Trust starts with clear boundaries.

No leaderboard. No developer worth score. No token spend scoreboard. Contribution.dev reviews the evidence, shows the limits, and leaves you in control.

Private by default

The CLI does not upload raw code, raw diffs, secrets, author emails, or private paths by default. Public exports fail closed when redaction cannot be trusted.

Product boundaries

  • No developer worth score
  • No leaderboard
  • No automatic raw code upload
  • No AI-code detector claims
  • No token ROI without attribution
  • No business ROI without outcome data

Run a readiness audit.

Find the five changes most likely to make agents faster, safer, and less wasteful in this repo.