Qneolabs

**Handoff Auditor — Evidence-Based AI Project Handoff Verification** Handoff Auditor is a standalone application developed by Qneolabs for the Meta Global AI Developer Hackathon 2026, under the Indie Developer Track and Build for Your World theme. AI-assisted development makes it easier to build software, but transferring work between developers, AI agents, and sessions introduces a critical problem: handoff documents may contain unsupported claims, outdated decisions, contradictory evidence, or incomplete approvals. Handoff Auditor addresses this by examining project handoff claims against supplied supporting evidence before those claims are carried forward. The application combines deterministic verification with semantic analysis using Meta's Muse Spark 1.3 model through Meta Model API. Its checks identify missing evidence, conflicting decisions, and superseded records. The Meta integration demonstrates how contextual contradictions can be detected and linked to exact source quotations. Users can inspect findings, record follow-up notes, and export evidence-linked continuation packets while keeping unresolved issues visible rather than treating them as completed decisions. The application runs locally. Its demonstration includes a saved response from a real Meta API test and separately labeled deterministic simulations. It does not perform live model inference during the video. The standalone Handoff Auditor implementation was developed during the hackathon using ChatGPT and Codex. Work Core and earlier concept planning predate the event and are not presented as new hackathon development.