aoci-spec

Aoci Code — Data skill for Claude Code

Data community

How to install Aoci Code

This entry records only its repository, not the path inside it, so there is no exact command to give. Open aoci-spec/aoci-code and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Aoci Code does

Local-first MCP server + CLI in Go: a persistent, governed codebase index that distills code and database schema knowledge into a map of the entire repository — code intelligence, context and memory that help AI coding agents (Claude Code, Codex, Cursor, opencode) understand complex systems faster and develop with greater precision.

Alternatives in Data

OpenViking — Self-evolving Context Database for AI Agents 33.3k ★
Opencode Mem — OpenCode plugin that gives coding agents persistent memory using local vector database 1.5k ★
Claude Code Memory Setup — Up to 71.5x fewer tokens per session on Claude Code with Obsidian + Graphify 947 ★

README

![AOCI-CODE logo — AI-Oriented Cognition Infrastructure](assets/aoci-logo-en.jpg)

AOCI-CODE

**AOCI-CODE is an indexing method that establishes global software-project cognition for AI Agents and provides a persistent, governable repository cognition map.**

🇺🇸 English | [🇨🇳 简体中文](README.zh-CN.md)

![Status](https://img.shields.io/badge/status-v0.1.0--rc14-orange) ![Runtime](https://img.shields.io/badge/runtime-local--first-blue) ![MCP](https://img.shields.io/badge/MCP-9%20tools-6f42c1) ![License](https://img.shields.io/badge/license-FSL--1.1--MIT-blue)

1. 🎯 What AOCI-CODE does

1.1 **Keep developing a system inside a Code Agent.** In Codex, Claude Code, Cursor, OpenCode, and similar Agents, grow a system from nothing into a complete 500,000-line codebase plus its database. The Agent always works with the whole-system cognition the AOCI index provides instead of searching and re-reading the entire codebase for every task. Non-professional developers can keep iterating on their own systems; professional developers can hand the whole system to an Agent for collaborative evolution and keep their attention on architecture and design decisions.

1.2 **Understand an existing system in one step and take over its development.** For an existing system of up to 500,000 lines, one sentence has the Agent build whole-system cognition; it reports how well it has mastered the system, then takes over development seamlessly. The supported scale is set by the index size, not by the line count as such: in practice a 700,000-line commercial system is developed this way today, with an index of about 300K tokens.

1.3 **Change the people, the Agent, or the conversation without losing cognition.** The index lives in the repository next to the code and is versioned by Git. When a project changes its project manager or developers, switches to another Agent, or opens a new conversation, one read of the index picks up where things left off, seamlessly across Agents and conversatio