Audit agent-rules files across seven dimensions and emit line-referenced findings.
npx skills add https://github.com/bpawlakj/ai-devkit --skill rule-review
Please help me install this Agent Skill. Skill: rule-review Source: https://github.com/bpawlakj/ai-devkit/tree/main/claude/skills/rule-review Command: npx skills add https://github.com/bpawlakj/ai-devkit --skill rule-review
This Skill helps you verify that an agent-rules file is correct, efficient, non-redundant, and actually reachable—so your agent follows the rules instead of ignoring or contradicting them.
Run a read-only audit of your rules file by instructing the AI:
“Review AGENTS.md for redundancy, drift, ordering, and dead rules.”
High-intent search queries and answers about installing and using this skill.
Cross-tool drift detection compares common sibling rule sources like AGENTS.md, CLAUDE.md, GitHub Copilot, and Cursor configurations to find contradictions for the same scope, ensuring your agent consistently follows intended constraints.
Cross-tool drift occurs when sibling rule sources like AGENTS.md and CLAUDE.md contain contradictions for the same scope. Detection compares these common files to find inconsistencies and emits a structured report with line-referenced findings.
Auditing agent-rules files checks seven dimensions: length, embedded code blocks, language precision, redundancy, ordering, cross-tool drift, and dead-rule reachability. This evidence-based approach ensures your agent follows constraints instead of ignoring them.
Agent-rules files are audited across seven dimensions: length, embedded code blocks, language precision, redundancy, ordering, cross-tool drift, and dead-rule reachability. Evidence-backed verdicts ensure agents follow intended constraints consistently.
Agent-rules auditing applies when authoring or maintaining AGENTS.md, CLAUDE.md, and related rule documents. Cross-tool drift detection compares common sibling rule sources including GitHub Copilot and Cursor to find scope contradictions.
A policy audit for prompt engineering rules covers seven dimensions: length, embedded code blocks, language precision, redundancy, ordering, cross-tool drift, and dead-rule reachability. Each dimension produces evidence-backed findings with line references.
pablontivAudit and optimize AI context files to reduce token waste.
potetoDetect and replace AI writing patterns in prose to produce a natural human voice.
rudironsoniGenerate AI coding tool configurations from a unified rule set.
ConductionNLAudit Nextcloud app files against openspec/app-config.json to detect drift.
