graph-explore

Query the code-review-graph MCP server to trace dependencies and map impact areas.

17|Updated Mar 30, 2026
One-click install
npx skills add https://github.com/alexiolan/craft-skills --skill graph-explore-alexiolan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: graph-explore
Source: https://github.com/alexiolan/craft-skills/tree/main/skills/graph-explore
Command: npx skills add https://github.com/alexiolan/craft-skills --skill graph-explore-alexiolan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Code repositories often suffer from unclear structure and unknown dependencies. Graph Explore provides a structured method to discover code relationships by dispatching a haiku agent to interact with the code-review-graph MCP server, generating an up-to-date map of domains, files, and connections.

Core Features & Use Cases

  • Semantic graph exploration: identify domain boundaries, key modules, and their interconnections.
  • Dependency tracing: map imports and dependents to understand blast radii and change impact.
  • Post-change review support: generate context for risk assessment and planning before implementing changes.

Quick Start

Describe the feature to explore and the target repository root to begin graph exploration with the MCP tools.

Frequently Asked Questions about graph-explore

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I trace dependency blast radius for planned code changes?

Code dependency blast radius tracing works by querying the code-review-graph MCP server to map imports and dependents. This identifies structural relationships and impact areas, generating up-to-date context for risk assessment before implementing planned changes.

What is semantic search for codebase exploration?

Semantic codebase exploration uses a haiku agent to query the code-review-graph MCP server, applying file summaries and embeddings. It identifies domain boundaries, key modules, and their interconnections to map structural relationships across repositories.

How do I map code graphs and domain boundaries in a repository?

Mapping code graphs involves dispatching a haiku agent to interact with the code-review-graph MCP server. The agent applies semantic search and dependency tracing to discover domain boundaries, files, and their interconnections within the target repository root.

Do I need a specific MCP server to analyze code review graphs?

Yes, code review graph analysis requires the code-review-graph MCP server. The Skill ensures graph freshness and embeddings readiness before performing exploration, impact analysis, or post-change review modes using that specific server connection.

Can I use semantic search to review code changes before implementation?

Yes, semantic search supports post-change code review by querying the code-review-graph MCP server to evaluate blast radius and impact areas. It generates context for risk assessment and planning before implementing modifications.

What are the limitations of graph exploration for risk analysis?

Graph exploration risk analysis relies on the code-review-graph MCP server and requires graph freshness and embeddings readiness. It operates within explicit explore, impact, and review modes, meaning analysis is bounded by the queried repository structure.