Writing accurate conventional commits from complex staged changes takes time and can produce inconsistent results. This skill inspects staged Git changes and recommends a concise, scoped commit message.

Local applications often fail because runtimes, dependencies, ports, variables, permissions, or services are misconfigured. Env Doctor checks these conditions and returns prioritized, platform-specific fixes without exposing environment values.

Verification tasks often expand beyond their acceptance criteria. This skill selects sufficient checks, reports exact outcomes, and stops when proof is complete.

Broad fixes can introduce unrelated regressions and overwrite existing work. This skill traces each bug to its responsible layer and verifies a focused correction.

Structural code changes can introduce subtle regressions and expand scope. This skill defines behavior boundaries and brackets focused refactors with verification.

Complex migrations can lose data or break compatibility during staged rollouts. This skill structures transitions around preservation, verification, rollback, and mixed-version safety.

Feature work often grows beyond acceptance needs and weakens architectural ownership. Lean Build defines a narrow outcome, reuses repository patterns, and stops after focused verification.

Premature edits can hide causes and introduce new failures. This skill organizes evidence, ranks hypotheses, and identifies a credible mechanism before implementation.

Los archivos de memoria largos consumen contexto y encarecen las sesiones de IA repetidas. Esta skill comprime archivos de memoria markdown y de texto, preservando detalles técnicos, código, URLs y estructura.

Los usuarios de Claude Code necesitan recibos fiables de tokens sin pedir al modelo que estime el uso. Esta skill documenta un comando basado en hooks que informa de los tokens de sesión y los ahorros estimados a partir del registro local.

LLM requests can be difficult to measure across SDKs and providers. This skill identifies live callsites, routes them through Caveman, and verifies one measured request.

Las revisiones de pull requests suelen volverse demasiado extensas para actuar con rapidez. Esta skill da formato a los comentarios como observaciones concisas, listas para pegar, con ubicación, problema y solución.

Aggregate repository profiles do not identify a safe or valuable code change by themselves. This skill tests one approved candidate against a controlled baseline.

Experiment lifecycle decisions can be unsafe when evidence or server safeguards are incomplete. This skill reviews current results and recommends one supported action without executing mutations.

Large configuration files and repeated context increase token usage. This skill reviews Caveman reports and applies measured changes with individual approval.

Los usuarios pueden olvidar los modos, activadores y opciones de configuración de Caveman. Esta skill muestra una tarjeta de referencia compacta sin cambiar el modo ni guardar estado.

Finding relevant code across an unfamiliar repository can consume time and context. This skill searches broadly and returns concise, verified file and line citations.

LLM cost and quality changes are difficult to explain from aggregates alone. This skill produces bounded, trace-supported Caveman Cloud reviews without changing experiments.

Unlabeled LLM traffic hides which jobs create spend. This skill discovers repository workflows, proposes durable labels, applies approved changes, and verifies attribution.

Los archivos grandes de memoria del proyecto consumen tokens cada vez que comienza una sesión de programación con IA. Esta skill comprime archivos de memoria en prosa, conservando los detalles técnicos y guardando una copia de seguridad legible.