cache-lint

Detect prompt cache efficiency violations in cursor/agents and cursor/skills files.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/cuongnm-dev/team-ai-config --skill cache-lint
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cache-lint
Source: https://github.com/cuongnm-dev/team-ai-config/tree/main/cursor/skills/cache-lint
Command: npx skills add https://github.com/cuongnm-dev/team-ai-config --skill cache-lint

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill hard-enforces frozen-header + 4-block invariants in agent and skill prompts to keep prompt cache predictable, reducing Claude compute and costs.

Core Features & Use Cases

  • Detects trailing whitespace, CRLF line endings, missing frontmatter, and other policy violations across cursor/agents and cursor/skills.
  • Enforces 4-block header order, FROZEN_HEADER safety, and kebab-case field naming to improve consistency and tooling reliability.
  • Useful during pre-commit checks, CI pipelines, or periodic audits to ensure prompt quality and stable deployments.

Quick Start

Audit your Cursor prompt configurations to identify violations and automatically fix whitespace and line endings.

Frequently Asked Questions about cache-lint

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

FAQPage Schema
How do I enforce prompt cache rules in Claude agent prompts?

Prompt cache rules are enforced by auditing files under cursor/agents and cursor/skills, checking frontmatter, header order, dynamic placeholders, and whitespace to report violations that affect cache efficiency.

Why does my prompt cache efficiency drop when using dynamic placeholders?

Prompt cache efficiency drops because dynamic placeholders disrupt frozen header invariants required for stable caching. This Skill identifies and reports placeholder violations in agent and skill prompts to restore predictable caching.

What is the best way to lint markdown frontmatter for CI pipelines?

Linting markdown frontmatter for CI pipelines involves running deterministic Python 3 checks to enforce kebab-case field naming and 4-block header order, ensuring prompt quality and stable deployments without external dependencies.

Does this prompt validation tool require external dependencies?

This prompt validation tool requires no external dependencies. It uses Python 3 with standard library for deterministic checks, allowing you to audit frontmatter and whitespace violations without installing additional packages.

Can I automatically fix whitespace and line endings in cursor skill files?

Yes, you can fix whitespace and line endings in cursor skill files. The Skill detects trailing whitespace and CRLF line endings during audits, automatically resolving these policy violations to maintain prompt cache stability.

When should I not use automated prompt linting for Claude compute?

Automated prompt linting for Claude compute should be avoided when markdown files lack a stable frozen header or 4-block structure, as the checks enforce these specific cache invariants rather than general markdown formatting rules.