Loading...
Loading...
Found 2,150 Skills
Validate and optimize CLAUDE.md files using Anthropic's best practices for focused sessions. Detects contradictions, redundancy, excessive length (200+ lines), emphasis overuse (2%+ density), broken links, and orphaned sections. Scores health 0-50 points. Suggests safe automated fixes and extraction opportunities. Use when editing CLAUDE.md, before commits, when document grows past 200 lines, user says "optimize CLAUDE.md", "check contradictions", "validate documentation", or during quarterly reviews. Works with project and global CLAUDE.md files (.md extension). Based on Anthropic 2025 context engineering best practices.
Prompt caching for Claude API to reduce latency by up to 85% and costs by up to 90%. Activate for cache_control, ephemeral caching, cache breakpoints, and performance optimization.
Use Document Mind (DocMind) via Node.js SDK to submit document parsing jobs and poll results. Designed for Claude Code/Codex document understanding workflows.
Add or change allowed commands in AI agent permission configs (OpenCode, Claude Code)
Validate Claude Code plugin structure, manifests, frontmatter, and dependencies. Use when user mentions "validate plugin", "check plugin health", "plugin broken", "plugin not loading", "plugin errors", "plugin.json invalid", "agent not working", "check plugin structure", or reports plugin-related issues.
You are updating this project's CLAUDE.md to reflect **meta learnings** (conventions, philosophy, gotchas) and **significant recent changes** to the codebase.
Comprehensive TUI/Web dashboard for Claude Code monitoring
Expert at integrating MCP servers into Claude Code plugins. Auto-invokes when configuring MCP servers (stdio/SSE/HTTP/WebSocket), writing .mcp.json files, or adding external tool integrations.
This skill should be used when the user asks to "create AGENTS.md", "update AGENTS.md", "maintain agent docs", "set up CLAUDE.md", or needs to keep agent instructions concise. Enforces research-backed best practices for minimal, high-signal agent documentation.
Automatically updates terminal window title to reflect the current high-level task. Use at the start of every Claude Code session when the user provides their first prompt, and whenever the user switches to a distinctly new high-level task. Helps developers manage multiple Claude Code terminals by providing clear, at-a-glance identification of what each terminal is working on.
Manually teach Claude Code an error pattern and its solution, storing it in the learning database with high confidence. Use when user provides an explicit "error -> solution" pair, wants to pre-load fix knowledge, or corrects a previous bad fix. Use for "/learn", "teach pattern", "remember this fix". Do NOT use for automatic error learning (that is the error-learner hook), debugging live issues, or querying existing patterns.
First-time Perses setup pipeline: discover or deploy server, configure MCP connection, create initial project, add datasources, and verify connectivity. 4-phase pipeline: DISCOVER, CONNECT, CONFIGURE, VALIDATE. Use when setting up Perses for the first time, connecting Claude Code to an existing Perses instance, or onboarding a new team to Perses. Use for "perses onboard", "setup perses", "connect to perses", "perses getting started". Do NOT use for dashboard creation (use perses-dashboard-create) or server deployment details (use perses-deploy).