Skip to main content

​
Installation

Get EchoVault running in under 2 minutes.

​
Install EchoVault

1

Install from GitHub

EchoVault requires Python 3.10 or higher.
pip install git+https://github.com/mraza007/echovault.git
This installs the memory CLI globally.
2

Initialize the memory vault

Create the vault directory where memories will be stored:
memory init
This creates ~/.memory/vault/ for storing Markdown session files and ~/.memory/index.db for the SQLite search index.
3

Configure your agent

Run the setup command for your agent. This installs the MCP server configuration automatically.
memory setup claude-code
By default, the config is installed globally. To install for a specific project instead:
cd ~/my-project
memory setup claude-code --project   # writes .mcp.json in project root
memory setup opencode --project      # writes opencode.json in project root
memory setup codex --project         # writes .codex/config.toml + AGENTS.md
That’s it. Your agent now has access to memory_save, memory_search, and memory_context tools via MCP.

​
Configure embeddings (optional)

Embeddings enable semantic search. Without them, you still get fast keyword search via FTS5.
1

Generate a starter config

memory config init
This creates ~/.memory/config.yaml with sensible defaults.
2

Edit the config

Open ~/.memory/config.yaml and configure your embedding provider:
embedding:
  provider: ollama
  model: nomic-embed-text
  # base_url: http://localhost:11434

enrichment:
  provider: none

context:
  semantic: auto
  topup_recent: true
3

Reindex existing memories (if any)

If you already have memories saved and want to add embeddings:
memory reindex
This rebuilds the vector index with your new embedding provider.
What each config section does:
  • embedding — How memories get turned into vectors for semantic search. ollama runs locally, openai calls cloud APIs.
  • enrichment — Optional LLM step that enhances memories before storing (better summaries, auto-tags). Set to none to skip.
  • context — Controls how memories are retrieved at session start. auto uses vector search when embeddings are available, falls back to keywords. topup_recent also includes recent memories.

​
Configure memory location (optional)

By default, EchoVault stores data in ~/.memory. You can change this in two ways:
  • MEMORY_HOME=/path/to/memory (highest priority, per-shell/per-process)
  • memory config set-home /path/to/memory (persistent default)
memory config set-home /path/to/memory
memory config clear-home
memory config
memory config shows both memory_home and memory_home_source (env, config, or default).

​
Verify installation

Check that everything is configured correctly: 
memory config
You should see your effective configuration, including memory_home and any embedding provider settings.

​
What gets installed

AgentSetup commandWhat gets installed
Claude Codememory setup claude-codeMCP server in .mcp.json (project) or ~/.claude.json (global)
Cursormemory setup cursorMCP server in .cursor/mcp.json
Codexmemory setup codexMCP server in .codex/config.toml + AGENTS.md fallback
OpenCodememory setup opencodeMCP server in opencode.json (project) or ~/.config/opencode/opencode.json (global)
All agents share the same memory vault at your effective memory_home path (default ~/.memory/).

​
Uninstall

To remove EchoVault from an agent: 
memory uninstall claude-code   # or: cursor, codex, opencode
To uninstall the package: 
pip uninstall echovault
To also remove all stored memories: 
rm -rf ~/.memory/

​
Next steps

Quick start

Save your first memory and see how agents use it

Build docs developers (and LLMs) love

Get started for freeTalk to us