What problem does it solve? Subagent results are injected verbatim into the main conversation context, so verbose agent output quickly exhausts the context window during long coding sessions. This Skill decides when to delegate work to caveman-compressed subagents so each delegation costs roughly 60% fewer tokens. ## Core Features & Use Cases - Delegation decision guide: Maps task types to the right agent — cavecrew-investigator for locating code, cavecrew-builder for 1-2 file edits, cavecrew-reviewer for diff audits — versus vanilla agents or the main thread. - Structured output contracts: Defines predictable, grep-friendly result formats (path:line findings, verified edit confirmations, severity-tagged review findings) the main thread can rely on. - Chaining patterns: Documents locate → fix → verify pipelines, parallel investigator scouting, and single-shot edits for known sites. - Use Case: During a long refactoring session, spawn cavecrew-investigator to find all callers of a function, hand the top site to cavecrew-builder for the edit, then have cavecrew-reviewer audit the diff — all while keeping main-context usage low. ## Quick Start Ask the agent to use cavecrew to locate where a function is defined, apply a small edit, and review the resulting diff with compressed subagent output.
