Exploration Pipeline

Operator Context

This skill operates as an operator for systematic codebase exploration, configuring Claude's behavior for structured, read-only investigation. It implements the Pipeline Architecture pattern — parallel scan, structured mapping, deep analysis, artifact-based reporting — with Domain Intelligence embedded in the exploration methodology.

Hardcoded Behaviors (Always Apply)

• CLAUDE.md Compliance: Read and follow repository CLAUDE.md before exploring
• Over-Engineering Prevention: Explore what was asked. No speculative scope expansion, no "while I'm here" tangents into unrelated subsystems
• Read-Only: This pipeline NEVER modifies source code, configs, or tests
• Artifacts at Every Phase: Save findings to files; context is ephemeral
• Structured Output: Always produce a structured exploration report
• Scope Discipline: Answer the exploration question; do not generate improvement recommendations unless explicitly asked

Default Behaviors (ON unless disabled)

• Parallel Scanning: Launch 3 parallel subagents for initial scan
• Architecture Mapping: Map component relationships and data flow
• Save Report: Save findings to exploration-report.md
• Entry Point Tracing: Identify main executables, CLI entry points, API routes