|
Coograph β graph-first context for AI coding agents
Open-source SQLite dependency graph. Cuts token use 68%, tool calls 2.75Γ β median across the committed benchmark. MIT. Works with Claude Code, Copilot, Code...
|
83 % |
|
CatnipCoder | Code intelligence, LLM-powered developer tools, and the JVM
Engineering deep-dives on code intelligence, LLM-powered developer tools, and the JVM ecosystem, by Vasilis Soumakis.
|
82 % |
|
SocratiCode Cloud | The codebase context engine your team shares
SocratiCode is a codebase context engine, not another AI. It indexes your repositories (code, dependency graphs, database schemas, API specs and architecture...
|
82 % |
|
Graphify β Save up to 80% of Claude Tokens
Graphify is an open-source AI token-cost optimizer for Claude. It answers from a compact knowledge graph of your code, docs, papers and diagrams instead of r...
|
82 % |
|
Nodalyze β Save up to 80% of Claude Tokens
Nodalyze is an open-source AI token-cost optimizer for Claude. It answers from a compact knowledge graph of your code, docs, papers and diagrams instead of r...
|
82 % |
|
Cambra
AI generates code faster than teams can validate it. Cambra closes the gap β native verification, application branching, and a unified model built for the pa...
|
81 % |
|
ragpilot β RAG MCP server for local codebases
ragpilot is a local RAG (Retrieval-Augmented Generation) MCP server that gives AI coding agents semantic search, symbol navigation, impact analysis, and toke...
|
81 % |
|
dekx β the support copilot that investigates
dekx is a copilot for support and solutions engineers. It investigates tickets against your code, docs and past tickets β locally, on your own machines.
|
81 % |
|
Graph Engineering | Evolve prompting into a system you can verify.
A hands-on course in agentic graph engineering: knowledge-graph memory, cheap extraction, selective retrieval, and routing policy. Driven with Hermes Agent a...
|
81 % |
|
Home | NeaByteLab
Software engineer crafting open-source tools, shipping developer products, and writing in-depth about systems, design, and the craft of building reliable sof...
|
80 % |
|
Coding Architect
I help AI coding agents work on codebases too big for them I build open-source tools β code knowledge graphs, code intelligence, and architecture tooling β t...
|
80 % |
|
noodlbox β context infrastructure for coding agents
The nbx CLI compiles your codebase into a local knowledge graph your agents can query. LLM-free.
|
80 % |
|
Code-Graph-RAG - AI-Powered Codebase Analysis
An AI-powered codebase analysis tool that builds knowledge graphs from multi-language codebases using Tree-sitter parsing.
|
80 % |
|
Depwire β The Missing Context Layer for AI Coding Assistants
Depwire gives AI coding assistants a deterministic dependency graph via MCP. 24 tools for impact analysis, architecture health, dead code detection, What If...
|
80 % |
|
Glass Slipper Β· The one-click hybrid LLM solution
Glass Slipper routes your AI coding tasks between frontier and local models automatically. Work in Claude, Codex, or Cursor. Heavy thinking goes to the cloud...
|
80 % |
|
CodeWax
An AI-powered RAG assistant for codebase exploration and analysis
|
80 % |
|
CodeHub β the vibe coding nervous system
The vibe coding nervous system: the world's public codebase as fresh, queryable code graphs β current APIs, real usage, version diffs, over MCP, at a fractio...
|
80 % |
|
Assembly β ground your coding agent in your design system
A self-hostable knowledge server that turns any design-system corpus into queryable knowledge for coding agents β hybrid vector + full-text retrieval over MC...
|
80 % |
|
Vellis β open-source context graph engine for AI agents Β· volantlabs.ai
Vellis is an Apache-licensed, open-source context graph engine for AI agents: model meaning explicitly, validate change, connect MCP-capable tools, and recov...
|
80 % |
|
Contextifly: Give your AI a memory of your codebase
A persistent context engine for AI coding assistants. Contextifly compiles your React, NestJS and Flutter code into a live, deterministic knowledge graph. 10...
|
80 % |
Detected social media networks