
from repowise1,249
Use Repowise tools to quickly get architecture overviews, dependency paths, and focused context for exploring a codebase indexed with Repowise.
This skill teaches the agent how to use Repowise's code-intelligence tools to explore and understand a repository that has been indexed by Repowise. Instead of scanning raw files, the skill recommends calling Repowise API-like helpers (get_overview, search_codebase, get_context, get_dependency_path, get_architecture_diagram) to obtain architecture summaries, module maps, ownership info, and mermaid diagrams. It includes error-handling guidance and fallbacks when Repowise lacks coverage.
Activate this skill when asked to explain high-level architecture, locate where a feature is implemented, trace dependencies between modules, or produce a visual subsystem diagram. It's intended as the first step in any analysis task on projects with a .repowise/ index — call get_overview(), then narrow with targeted searches and context requests.
Best used by agents integrated with Repowise/Claude-Code tooling or other agents that can call the project's Repowise helpers. Works well in Claude/Code or any environment where the Repowise binary or APIs are available.
Codebase Exploration (Repowise) is a documentation-only skill with no bundled scripts. It guides agents to use Repowise MCP tools for architecture overviews, codebase search, dependency path analysis, and visual diagrams. Clean SKILL.md with clear triggers, well-structured workflow sections, and practical error handling guidance. No security concerns whatsoever — purely instructional with no executable content.
Niche but well-crafted skill. Documentation-only, no scripts. Would benefit from a references/ section with example outputs. The user-invocable: false flag means it only activates when a .repowise/ directory is detected — good guard.