Can you independently reproduce PactVerity’s evidence benchmark?

We have opened a public independent-evaluation programme for developers, security reviewers and technical writers.

The objective is not an endorsement. Choose one claim, publish the exact environment, command, public input source, complete output and limitations, and disclose any material connection.

Available routes:

- reproduce the three-scenario benchmark;

- review the receipt evidence model;

- test a JavaScript, Python or MCP reference integration;

- cite or critique the open dataset.

No token purchase, wallet connection or positive conclusion is required. Verified external runs remain zero until an evidence package can be reproduced.

Evaluation programme: https://pactverity.com/independent-evaluation

Repository: https://github.com/PactVerityHQ/pactverity

Research kit: https://pactverity.com/research