styler-ai/ProjectAtlas

Every file not opened. Every folder not explored. Tokens saved. A local repository map for coding agents: purposes, source, symbols, and dependencies. Native Rust CLI and MCP, isolated worktrees, and subagent hooks.

What it solves

ProjectAtlas addresses the high token costs and inefficiency of AI coding agents when navigating large codebases. By providing agents with a structured, intelligent map of a repository, it prevents them from reading irrelevant files and helps them locate the correct code faster, significantly reducing token consumption.

How it works

It functions as a native Rust CLI tool and MCP (Model Context Protocol) server that builds a persistent, local SQLite-based index of your project. It scans your repository—respecting .gitignore—to create a graph of files, folders, symbols, and relationships. When an agent is working, it queries this local database to get summaries, file outlines, and specific code slices rather than blindly scanning the entire file system.

Who it’s for

It is designed for developers who use AI coding agents like Claude Code, Codex, or OpenCode and want to improve the agent's performance, accuracy, and cost-efficiency when working on large, complex software projects.

Highlights

  • Significant Token Reduction: Can reduce token usage by over 90% by guiding agents directly to relevant code.
  • Local-First Architecture: Stores all intelligence in a local .projectatlas/ directory; no external cloud indexing or credentials required.
  • Agent-Ready: Provides native MCP support, allowing agents to perform task-oriented discovery, symbol lookups, and health checks.
  • Worktree Support: Seamlessly handles multiple Git worktrees, allowing agents to maintain context across different branches or versions of a project.
  • Performance-Focused: Built in Rust with a SQLite backend for fast, reliable repository intelligence.

Related

  • Project
    yc-duan/fastctx

    FastCtx is a framework that lets LLM agents like Codex and Claude Code run across multiple personal machines as a unified “World,” sharing file‑system access and a common action journal while respecting each provider’s usage terms.

  • Project
    DeusData/codebase-memory-mcp
  • Project
    bgauryy/octocode

    An agentic code research platform that provides AI agents and developers with token-efficient, evidence-based context from local and GitHub repositories using a Rust-powered engine.

  • Project
    aovestdipaperino/tokensave

    tokensave is a Rust‑based local server that builds a semantic knowledge graph of a codebase, letting AI coding agents query symbols, relationships, and snippets in a single call. By replacing costly grep/glob scans, it cuts token usage, supports multi‑branch indexing, offers crash‑resilient syncing, and integrates with dozens of LLM‑powered coding tools.

  • Project
    syncable-dev/memtrace-public

    A structural memory system for AI coding agents that turns codebases into local, bi-temporal knowledge graphs for fast, low-cost architectural queries.