I Built BountyPilot: An Open-Weight AI That Helps Developers Find Coding Opportunities Worth Their Time

Description and use

BountyPilot helps independent developers quickly assess whether to tackle issues with promised bounties by providing verified reward evidence, skill requirements, difficulty estimates, and risk flags.

Problem

Independent developers struggle to decide whether a GitHub issue with a bounty title actually offers a paid opportunity, due to unverified reward claims and scattered information.

Outcome

BountyPilot enables developers to quickly assess whether a bounty issue is worth pursuing by providing verified reward evidence, skill alignment, difficulty estimates, and risk flags, reducing time spent on unverified leads.

Workflow

• Search GitHub issues via REST API
• Retrieve issue source and discussion excerpts
• Run local Gemma 3 4B inference through Ollama
• Validate model output schema and citations
• Present analysis with reward evidence and risk notes
• Save opportunity to local SQLite pipeline
• Reload or revisit saved opportunities

Evidence

It searches real public GitHub issues, retrieves their sources, and puts reward evidence beside the technical task.

Source

The model must return structured JSON. The application checks the schema, source identifiers, unique citation identifiers, exact quote membership and risk-to-citation references.

Source

Real-issue Gemma 4B analysis 11.394 seconds on an M4 Pro Mac with 48 GiB RAM; one observed request

Source