Skip to content

AI Rules Linter

Lint CLAUDE.md, .cursorrules, and copilot-instructions files for redundancy, conflicting instructions, missing sections, and token efficiency.

AI Rules Linter

Score 82, 5 findings

Format
CLAUDE.md
Score
82
Tokens
113
Findings
5

Score breakdown

Completeness30%
75
Specificity25%
56
Efficiency25%
100
Consistency20%
100
Cost per conversation message (input)
GPT-6 Astra: $0.0011Claude Fable 5.1: $0.0011Gemini 3.1 Pro Preview: $0.0002

Sections (4)

SectionTokensLines%
Identity183-5
16%
Code Style316-10
27%
Testing1611-14
14%
Error Handling3015-18
27%

Present (6/8)

✓ Identity / Role Definition
✓ Code Style Preferences
✓ Testing Requirements
✓ Error Handling Preferences
✓ Forbidden Patterns
✓ Language / Framework Preferences

Missing (2/8)

✗ File / Project Structure
✗ Output Format Preferences

Findings (5)

infomissing

Missing: File / Project Structure

No section addressing file / project structure was detected.

Suggestion: Add a section with clear directives for file / project structure.
infomissing

Missing: Output Format Preferences

No section addressing output format preferences was detected.

Suggestion: Add a section with clear directives for output format preferences.
warningvague

Vague phrasing: "Try to"

Found 1 instance of "try to". Vague instructions reduce AI compliance.

Suggestion: Remove "try to" and state the rule directly, e.g. "keep functions under 50 lines"
warningvague

Vague phrasing: "when appropriate"

Found 2 instances of "when appropriate". Vague instructions reduce AI compliance.

Suggestion: Define the specific conditions, e.g. "add error handling for all async operations"
warningvague

Vague phrasing: "Consider"

Found 1 instance of "consider". Vague instructions reduce AI compliance.

Suggestion: Be directive: "use X" instead of "consider using X"

Updated . Provided as is. Check the output before you rely on it in production.

How to use AI Rules Linter

  1. 1

    Paste your rules file

    Enter the contents of your CLAUDE.md, .cursorrules, copilot-instructions.md, or .windsurfrules file. The tool auto-detects the format.

  2. 2

    Review the effectiveness score

    Check the overall score (0-100) based on completeness, specificity, efficiency, and consistency. The section breakdown shows token usage per heading.

  3. 3

    Address findings

    Review each finding — redundant instructions, conflicting directives, missing sections, or vague language — and apply the specific suggestions provided.

  4. 4

    Optimize token usage

    The token breakdown shows where your budget is spent. Reduce high-token sections that provide little value and ensure critical sections have adequate coverage.

Questions and answers

What is AI Rules Linter?
AI rules files such as CLAUDE.md, .cursorrules, .windsurfrules and copilot-instructions.md are standing instructions a coding assistant reads on every request. This linter flags redundant, conflicting, missing and vague rules and scores the file 0-100 on completeness, specificity, token efficiency and consistency.
What does the effectiveness score measure?
The score (0-100) weights four factors: section completeness (30%), instruction specificity (25%), token efficiency (25%), and internal consistency (20%). Higher scores indicate rule files that will produce better AI assistant behavior.
Does it send my rules file to a server?
No. All linting happens entirely in your browser. Your configuration files never leave your device.
Which rule file formats are supported?
CLAUDE.md, .cursorrules, copilot-instructions.md, and .windsurfrules. The tool auto-detects the format from content patterns and applies format-specific checks.
For AI agents: how to call this tool

Machine-readable contract, endpoints and examples. Humans can ignore this section.

Best Path For Builders

Browser workflow

Runs instantly in the browser with private local processing and copy/export-ready output.

Browser Workflow

This tool is optimized for instant in-browser execution with local data handling. Run it here and copy/export the output directly.

/ai-rules-linter/

For automation planning, fetch the canonical contract at /api/tool/ai-rules-linter.json.