What problem does it solve? Agents and engineers often make blind changes in unfamiliar or large codebases, causing regressions in distant files and violating project conventions. This Skill performs read-only reconnaissance before any code is written, producing a structured context map that downstream agents consume. ## Core Features & Use Cases - WHY/WHAT/HOW Context Mapping: Captures intent, scope, dependencies, data flows, and an execution strategy as structured YAML with file paths and line numbers as evidence. - Three Mapping Modes: Full mapping for unfamiliar codebases, incremental mapping for targeted changes, and conflict detection when multiple agents touch shared files. - Typed Capability Contracts: Five capabilities (directory scan, dependency analysis, convention detection, change impact analysis, conflict detection) with defined inputs, outputs, and post-execution checks. - Use Case: Before refactoring authentication across 23 routes, run a full mapping to trace the token validation data flow, identify all downstream consumers of the auth middleware, and hand the resulting context map to implementation agents. ## Quick Start Ask the agent to map the context of your codebase for a planned change, for example: map this repository and identify every file affected by modifying the auth handler before I start the refactor.
