~97% token reduction for AI coding sessions — zero deps, 38 languages, MCP server
-
Updated
Nov 21, 2026 - JavaScript
~97% token reduction for AI coding sessions — zero deps, 38 languages, MCP server
A lightweight CPU only memory approach with ranked retrieval. Simple, yet effective.
Local code search for AI agents: cheaper, faster and sweeter. Because maybe grep isn't all you need... 🍬
goto.dev - Pull request superpowers
Local-first repo memory and context for coding agents.
Public distribution surface for jarvis — local-first code intelligence MCP server (SCIP navigation + Zoekt search). Installer, Claude Code / Codex / Cursor plugins, and prebuilt binaries.
Constellation Extension for Gemini CLI
AI-powered search capabilities for AI assistants using the Tavily API and Model Context Protocol (MCP)
Persistent engineering memory for coding agents. A DeepSeek Harness plugin that captures what the agent learned, keeps only what held up, and recalls it automatically. Local SQLite, no daemon, no API keys.
A DeepSeek Harness plugin that puts graft — a prebuilt graph of every symbol, its file:line span, and who calls what — in front of both you and the model.
The truth layer for coding agents: labelled caller answers and a stop check. Plugin and setup files; engine closed.
CodeGraphX (CGX) — a local, token-efficient codebase graph engine & MCP server for AI coding agents. Tree-sitter parsing, a bi-temporal SQLite semantic graph, O(1) symbol lookup, and impact/blast-radius tracing so agents answer 'what breaks if I change this?' in a few hundred tokens.
DeepSeek Harness plugin that gives coding agents a real code graph: six MCP tools for symbol search, AST snippets, call-path tracing, architecture analysis, and an auto-started graph UI.
AI-powered code search MCP server using Exa API for intelligent code search and retrieval in AI assistants
The public home for FifthDev and Fabric: reliable agent context, local code intelligence, and governed workflows.
A curated list of tools and resources for understanding codebases, dependencies, architecture, changes, and AI context.
AI-native codebase analysis engine for automated repository understanding, service maps, API schemas, and business logic extraction.
Pi-agent wrapper around codebase-memory-mcp: derived code knowledge (call graphs, blast radius) over a harness-owned trust boundary
Self-hosted code maps and MCP access for AI, with GitLab, GitHub and generic Git sync, a web dashboard, and repository-scoped tokens.
CodeMap: 把代码仓库编译成带 path:line 源码引用的代码地图 — DeepSeek Harness 插件,零模型调用、离线可跑、引用可校验。
To associate your repository with the code-intelligence topic, visit your repo's landing page and select "manage topics."