Browse Skills

Caveman Setup wires the current repository through the Caveman Cloud gateway so every LLM request is measured — cost, tokens, latency — with zero behavior change. Use when the user wants LLM spend observability added to an app or says 'set up caveman'.

Lean Build guides feature work with high overbuilding risk, enforcing repository reuse, strict scope, and an explicit stop condition. Use it for new behavior, product slices, or integrations where a minimal, complete path matters.

Learn Codebase systematically reads every source file in full to build a deep understanding of a new or unfamiliar project. Use it when asked to learn, read, prime, or get up to speed on a codebase.

Make Plan creates a detailed, phased implementation plan with documentation discovery agents that verify APIs and patterns before implementation. Use it when asked to plan a feature, task, or multi-step implementation, especially before executing with do.

Memory Search queries a persistent cross-session memory database so Claude can retrieve past work. Use it when users ask about previous sessions, such as "How did we solve X last time?"

Mode Creator interactively creates, installs, and verifies custom claude-mem modes with domain-specific observation types, concept tags, and optional Telegram alerts. Use it when customizing what claude-mem remembers for a project or workflow.

Pathfinder maps a codebase into feature-grouped flowcharts, identifies duplicated concerns across features, and proposes a unified architecture. Use it before a refactor or when asked to find the ideal path or unify duplicated systems.

Smart Explore performs token-optimized structural code search using tree-sitter AST parsing, allowing agents to understand code structure, find functions, and explore codebases without reading full files.

Do Plan executes a phased implementation plan by deploying subagents for each phase, verification, anti-pattern checks, code review, and only committing after verification passes. Use it when asked to execute, run, or carry out a plan, especially one created by make-plan.

How It Works explains how claude-mem captures observations, when memory injection kicks in, and where data lives. Use when the user asks how claude-mem works or what it is doing.

Knowledge Agent builds and queries AI-powered knowledge bases from claude-mem observations. Use it to create focused 'brains' from observation history, ask questions about past work patterns, or compile expertise on specific topics.

Make Plan creates a step-by-step plan for a given task or goal, using stored context from the memory engine to inform priorities and actions. Use it when breaking down complex requests into manageable, ordered steps.