review-rules

Reviews current work against task-instruction rule files and reports pass, warn, or fail compliance.

3|Updated Feb 15, 2025
One-click install
npx skills add https://github.com/mark-torres10/ai_tools --skill review-rules-mark-torres10
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: review-rules
Source: https://github.com/mark-torres10/ai_tools/tree/main/skills/review-rules
Command: npx skills add https://github.com/mark-torres10/ai_tools --skill review-rules-mark-torres10

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams define coding and workflow standards in rule files, but verifying that current work actually follows them is manual and inconsistent. This Skill runs a structured compliance review against the rules in agents/task_instructions/rules and produces an evidence-based pass/warn/fail assessment. ## Core Features & Use Cases - Rule Discovery and Selection: Resolves the ai_tools root, reads router.md when present, and selects only task-relevant rules based on frontmatter fields like when_to_use, scope, and applies_to. - Structured Compliance Matrix: Evaluates each selected rule against the current workspace context and records pass, warn, or fail with rationale. - Actionable Remediation: For every warn or fail, provides a concrete fix, a ranked list of top violations by impact, and a verification plan with tests or commands. - Use Case: After finishing a feature branch, ask for a rules-based review to confirm the changes comply with project standards before opening a pull request. ## Quick Start Review my current changes against the project task-instruction rules and give me a compliance matrix with fixes for any violations.

Frequently Asked Questions about review-rules

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check my code against project rules before submitting?

Run a rules-based compliance review that reads the rule files in agents/task_instructions/rules and evaluates your current changes against each one. The output is a compliance matrix with pass, warn, or fail status plus concrete remediation steps for any violations.

How does the review-rules skill select which rules to apply?

It first uses router.md if that file exists, then honors any user-specified rule filter. Otherwise it infers relevance from rule frontmatter fields such as when_to_use, scope, and applies_to.task_types, preferring task-relevant rules over loading everything.

What happens if the rules directory is not found?

The skill first checks for an ai_tools directory in the workspace root, then falls back to a canonical path at /Users/mark/Documents/projects/ai_tools/. If neither contains agents/task_instructions/rules, it asks the user for the rules directory and stops.

What output does a rules compliance review produce?

The review returns four sections: the rule set used with selection rationale, a compliance matrix mapping each rule to pass, warn, or fail with rationale, top violations ranked by impact with exact fixes, and a verification plan of tests or commands.

When should I not use a rules-based compliance review?

Avoid it when no rule files exist for your task type or when you need general code quality feedback rather than standards compliance. The review only evaluates against explicit rules and does not replace broader code review or testing.