Osloq: An AI agent that reproduces GitHub issues for you

This is the unglamorous but genuinely useful end of agentic coding, not writing more code, but verifying whether a bug report is even true before anyone touches it. It is exactly the kind of narrow, evidence backed agent that survives past the demo because it answers a yes or no question instead of promising to fix everything. Built and shipped solo, it is a good signal that the bar for a one person AI tooling company keeps dropping.