graph-intelligence

Analyze codebases through two-layer knowledge graphs combining AST call graphs and multimodal document graphs.

Updated Sep 2, 2026
One-click install
npx skills add https://github.com/nbiish/tts-mcp --skill graph-intelligence-nbiish
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: graph-intelligence
Source: https://github.com/nbiish/tts-mcp/tree/main/.agents/skills/graph-intelligence
Command: npx skills add https://github.com/nbiish/tts-mcp --skill graph-intelligence-nbiish

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Answering questions like "what calls this function", "what breaks if I change this", or "how does this module work" requires tracing call chains, dependencies, and documentation across large codebases, which is slow and error-prone when done by manual search. ## Core Features & Use Cases - GitNexus operational layer: AST-precise code graph with 15 MCP tools for symbol context, blast-radius impact analysis, multi-file rename, route mapping, and cross-repo contract bridging across 14 languages. - Graphify semantic layer: Multimodal graph covering code, docs, PDFs, images, and video, with PR triage, merge-conflict risk detection, god-node analysis, and exports to HTML, SVG, GraphML, Neo4j, and Obsidian. - Use Case: Before renaming a shared utility function, run an impact query to see every upstream caller grouped by depth and confidence, then check the PR dashboard to confirm no open pull request touches the same graph communities. ## Quick Start Ask the assistant to show the blast radius and callers of a specific function in the current repository using the graph intelligence tools.

Frequently Asked Questions about graph-intelligence

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

FAQPage Schema
How do I find what breaks if I change a function?▼

Use the GitNexus impact tool with the target symbol and direction set to upstream or downstream. It returns affected callers grouped by depth with confidence scores and risk levels, so you can review HIGH or CRITICAL impacts before editing.

What is the difference between GitNexus and Graphify?▼

GitNexus graphs code only, with AST-precise cross-file call resolution across 14 languages and pre-structured MCP tools. Graphify graphs code plus docs, PDFs, images, and video, and adds PR triage, but lacks cross-file call resolution.

Does Graphify send my code to external APIs?▼

Code files are parsed locally with tree-sitter and video is transcribed locally with faster-whisper. Only docs, PDFs, and images go to an LLM backend, and you can force fully local processing with the ollama backend.

Can I use knowledge graphs across multiple repositories?▼

GitNexus supports multi-repo groups natively through a registry and group-scoped queries with a contract registry for cross-repo routes. Graphify offers manual merging via merge-graphs or a global graph file for cross-project queries.

Why does my graph query return stale results?▼

The index reflects the last analyzed commit, so re-run gitnexus analyze or graphify update after pulling changes. The repos resource and status commands report staleness hints so you can check before querying.