claude-code-memory-setup
Up to 71.5x fewer tokens per session on Claude Code with Obsidian + Graphify. Persistent memory, codebase knowledge graphs, and chat import pipeline. 🇧🇷 PT-BR included.
Claude Code Memory Setup is a comprehensive agent skill configuration that adds persistent memory and codebase knowledge graphs to Claude Code sessions, reducing token consumption by up to 71.5x. It integrates Obsidian as a Zettelkasten-based second brain for cross-session context retention, Graphify for AST-based codebase mapping, and a chat import pipeline to preserve conversation insights. Developers configure custom slash commands (/resume, /save) and structured vault workflows that eliminate repetitive project re-explanation and file re-reading across sessions.
Key Features
Use Cases
- 01Developers working on long-term projects who need Claude Code to remember architecture decisions and conventions between sessions
- 02Teams maintaining large codebases (40+ files) who want to eliminate repetitive token costs from file re-reading
- 03Software engineers running multiple daily Claude Code sessions who need seamless context continuity
- 04Programmers managing multiple projects who want centralized cross-project knowledge linking in a single vault
- 05Developers who want to preserve and search through valuable insights from past Claude conversations
- 06Solo developers seeking a cost-effective alternative to enterprise AI memory solutions using free, open-source tools
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
claude-code-memory-setup — FAQ
What is Claude Code Memory Setup?+
It is an agent skill configuration guide that integrates Obsidian, Graphify, and custom slash commands into Claude Code to provide persistent memory across sessions and codebase knowledge graphs, reducing token usage by up to 71.5x.
How do I install this agent skill?+
Create an Obsidian vault, add a CLAUDE.md file at the vault root with the provided instructions, install Graphify via 'pip install graphifyy' and 'graphify install --platform claude', then set up the chat import pipeline with the provided Python scripts. Optionally configure the auto-save session hook in ~/.claude/settings.json.
Which AI clients does this work with?+
This setup is specifically designed for Claude Code (Anthropic's coding agent). The Graphify component also supports Cursor, Codex, OpenCode, and other platforms via 'graphify install --platform <name>'.
Do I need API keys or paid services?+
Obsidian is free, and Graphify's AST mode (default) costs zero tokens. Graphify's --mode deep semantic extraction requires an ANTHROPIC_API_KEY or MOONSHOT_API_KEY, but this is optional. The claude-conversation-extractor pip package is free.
How does the token reduction work?+
Instead of re-reading all project files every session (~20k tokens for 40 files), Claude Code queries a persistent Graphify knowledge graph (~280 tokens) and reads Obsidian vault context via /resume. The graph is built once locally using AST parsing with no ongoing token cost.
Can I use this with multiple projects?+
Yes, the guide recommends a single centralized Obsidian vault with project-specific subfolders. Each project gets its own architecture notes and Graphify subfolder, enabling cross-project knowledge linking and consistent memory across all work.
How do I install claude-code-memory-setup?+
Open the source repository on GitHub and follow its README. claude-code-memory-setup is a skill — MCP Agents Market links you directly to the official repo.
Is claude-code-memory-setup free?+
claude-code-memory-setup is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.