Autonomous bounty agent

Find it. Fix it. Get Paid.

The autonomous bounty agent for open source developers.

live-agent.log

> Connect GitHub profile to build a repository map

> Run web discovery to scrape public bounty sources

> Claim an issue to start diagnosis, fix, tests, and PR packaging

> Live agent events will stream here

0 bounties tracked

$0 open value

0 PR packages

How it works

The bounty loop, compressed.

From signal to merged PR, every step is visible, reversible, and tuned for open source contribution norms.

Scrape paid and free issue sources, then rank by profile fit and payout.

Use scanned repositories, languages, labels, and issue context to estimate fit.

Generate narrow patches that match the project conventions.

Detect the test stack, add regression coverage, and simulate CI.

Open a clean PR package and track outcome, receipt, and learning.

Features

Discovery, execution, proof, and learning all live in one interface.

Socket alerts for newly discovered paid and free issue matches.

Bounties are scored against your scanned language and domain fingerprint.

Branch, commit, PR body, and testing notes are generated for review.

Test commands are detected and CI output is captured for retry loops.

Merged and rejected PR outcomes can update ranking weights.

Completed cycles can export JSON receipts from real run data.

Pricing

No SaaS tax on your upside. Use your own model account and keep control of spend.

~$0.22

Estimated cost per full diagnosis, fix, test, and PR-description run.

BYO key

Bring your own OpenAI key. BountyHunter only orchestrates the workflow.

Included in every cycle