ki-audit

What problem does it solve?

The KI-Audit Skill uncovers typical AI-generated mistakes that block development and operations, such as hallucinated APIs, outdated syntax, incorrect nesting, wrong mappings, and semantic logic errors across configuration, dashboards, and code. It helps teams detect issues that are syntactically valid but semantically incorrect or that were introduced by AI-assisted edits, reducing time spent diagnosing subtle integration and configuration faults.

Core Features & Use Cases

• Cross-file / cross-stack checks: Grep and glob-based scans combined with referential checks against project reference documents to verify APIs, topics, and config keys.
• Targeted rule catalog: Applies a comprehensive catalog of checks (YAML/JSON formatting, PromQL/LogQL pitfalls, Docker compose keys, Grafana panel structure, Python async/import patterns, ESP32 constraints) and maps findings to severity and remediation.
• Structured audit reporting: Produces a standardized Markdown report with context, references used, categorized findings by ID, and prioritized recommendations. Use case: audit Grafana dashboards and Docker compose files for hallucinated settings and threshold logic before deploying monitoring to production.

Quick Start

Request an audit for the Grafana dashboards and Docker compose files to identify AI-generated configuration and logic errors.