What this app does
AI Delivery Evidence Checker evaluates user-supplied AI coding delivery materials and returns a structured evidence-sufficiency judgment. Results may identify supported, partially supported, unsupported, or not checked evidence states.
How this app helps
This app extracts verifiable delivery criteria from supplied AI coding requirements and assesses whether supplied delivery materials and test evidence support an AI completion claim. It can review individual evidence types or combine available results into a final delivery verdict.
What input it accepts
The app accepts text and structured fields provided by the user for the current review request, including requirements, criteria, completion claims, delivery evidence, test evidence, and verdict inputs.
What output it returns
The app returns structured results that classify evidence coverage, identify gaps, list warnings, and provide conservative verdict information based only on the supplied materials.
Available tools
extract_delivery_criteriacheck_delivery_evidencecheck_test_evidencegenerate_delivery_verdict
MCP endpoint
POST /mcp
What this app does NOT do
The app does not connect to GitHub, inspect complete repositories, modify code, run user project tests, create commits, push branches, create pull requests, deploy user applications, approve releases, or guarantee software correctness or absence of bugs.
Data handling
The app evaluates the materials supplied for the current request. Users should not submit passwords, API keys, tokens, private keys, or other secrets.
Support
Contact sidcraigau@gmail.com for support.