Urgent.News

What's breaking now, across thousands of outlets.

AI

NoWreck Deterministic AI Verifier — v0.13.0

NoWreck v0.13.0 — CI/CD Integration AstralXVoid / NoWreck A CLI tool that verifies AI coding assistant claims against actual structural changes — catching hallucinated functions, fake calls, and missed modifications before they ship. NoWreck Deterministic AI Verifier — v0.13.0 NoWreck is a deterministic structural verifier for AI-generated code-change claims. When an AI describes a code change,…

NoWreck v0.13.0, a deterministic AI verifier for AI-generated code-change claims, has been released. This CLI tool checks against structural changes in repositories to catch issues like hallucinated functions, fake calls, and missed modifications before code is shipped. The updated version includes several enhancements and new features.

Key updates in v0.13.0 include:

- SARIF output support for GitHub Code Scanning, SonarQube, and CodeQL

- New --output flag to write output to a file instead of standard output

- --format flag for selecting output format (json, sarif, junit)

- Automated git comparison for pull request verification using --compare flag

The tool supports various reference types for comparison, such as commit hash, branch name, tag, or HEAD~N syntax. It uses git archive for clean extraction of code before and after changes, and automatic cleanup via tempfile.TemporaryDirectory.

NoWreck v0.13.0 maintains backward compatibility with the --json flag, although a warning is issued. The 15 rules defined in the tool cover different claim types, each with its own rule for granular filtering. The CONFIRMED results are excluded by default in SARIF output, as SARIF is designed for problem identification.

This release builds upon the previous version, v0.12.0, which introduced CI/CD integration with machine-readable output formats (SARIF, JUnit) and automated git comparison for pull request verification.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

More from Wednesday 2 September →