Generate scenarios, execute coding agents, and report compliance rates.
npx skills add https://github.com/zero3041/PREP --skill skill-comply-zero3041
Please help me install this Agent Skill. Skill: skill-comply Source: https://github.com/zero3041/PREP/tree/main/.claude/skills/skills/skill-comply Command: npx skills add https://github.com/zero3041/PREP --skill skill-comply-zero3041
This Skill helps ensure that coding agents adhere to predefined skills, rules, and definitions, by automatically generating scenarios, executing them, and reporting compliance rates.
Run the skill-comply command on the path to the rule or skill file you want to test.
High-intent search queries and answers about installing and using this skill.
Measuring coding agent compliance involves automatically generating scenarios from .md files, executing them, and reporting compliance rates against predefined rules. It classifies tool call traces using LLM to verify adherence.
Automating code review for AI agents involves capturing tool call traces during scenario execution and classifying them against expected steps. This generates detailed reports with compliance scores and recommendations for process improvement.
Yes, measuring compliance requires Claude Code and Python libraries such as pyyaml, pytest, and subprocess for parsing and executing scenarios. These dependencies are essential for running the automated tests and generating reports.
The best way to verify agent adherence to new coding skills is to run automated scenario generation at supportive, neutral, and competing strictness levels. This tests whether the agent correctly follows the newly defined rules.
Yes, you can generate automated testing scenarios directly from .md files. The tool parses markdown definitions and automatically creates scenarios at three strictness levels: supportive, neutral, and competing.
You report compliance scores by executing scenarios, capturing tool call timelines, and using an LLM to classify traces against expected steps. This generates detailed reports with recommendations for promoting steps to hooks.
gaebalaiAutomate CAN bootstrapping for Raspberry Pi vehicle diagnostics environments.
idiaz01Generate behavioral specs and compliance reports from skill and rule files.
rahmatullahbossExecute predefined YAML workflow configurations to automate testing steps.
MatthieuszValidate TypeScript code against predefined coding standards using static analysis tools.
