engram · tstockham96 — Claude Code MCP server: features and setup
Stores project knowledge locally and helps an agent retrieve it by meaning. Useful when context needs to carry across sessions and projects.
claude mcp add engram -e GEMINI_API_KEY=<GEMINI_API_KEY> -- npx -y engram-sdkWhat engram does
engram saves memories in SQLite, links them into a knowledge graph, and retrieves relevant details through semantic search. An agent can ask about stored information, request a session briefing, or surface memories related to its current context. Import tools are available for Obsidian vaults and Claude Code memory.
An LLM helps extract and combine stored information, including spotting contradictions. Processing through Gemini requires your own GEMINI_API_KEY; the key is used for embeddings and memory handling. The data stays local, but Gemini-backed processing depends on that key.
To connect it to Claude Code, use claude mcp add. The setup below runs npx -y engram-sdk; another supported configuration appears in the FAQ.
What it helps with
- carry project context between sessions
- retrieve related details by meaning instead of text matching
- prepare a briefing before an agent session
- import memories from Obsidian or Claude Code
- find contradictions in stored knowledge
Example prompts for the agent
- Save that this project requires SQLite compatibility and local data storage.
- Before the session, summarize recent decisions and unfinished tasks.
- Import my Obsidian notes into the agent's memory.
How to install engram
Platform CLI
claude mcp add engram -e GEMINI_API_KEY=<GEMINI_API_KEY> -- npx -y engram-sdkReplace <…> with your own values
Platform CLI
codex mcp add tstockham96-engram --env GEMINI_API_KEY=<GEMINI_API_KEY> -- npx -y engram-sdkReplace <…> with your own values
Config file
{
"mcpServers": {
"tstockham96-engram": {
"command": "npx",
"args": [
"-y",
"engram-sdk"
],
"env": {
"GEMINI_API_KEY": "<GEMINI_API_KEY>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"tstockham96-engram": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"engram-sdk"
],
"env": {
"GEMINI_API_KEY": "<GEMINI_API_KEY>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e GEMINI_API_KEY=<GEMINI_API_KEY> tstockham96-engram npx -y engram-sdkReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about engram
How do I install engram in Claude Code?
Run claude mcp add engram -e GEMINI_API_KEY=<GEMINI_API_KEY> -- npx -y engram-sdk. Replace the placeholder with your Gemini key.
What does engram do?
It stores information in a local SQLite vault, connects memories in a knowledge graph, and retrieves relevant context through semantic search. It also supports session briefings and memory consolidation.
Does engram need an API key?
Gemini processing for embeddings and memory handling requires your own GEMINI_API_KEY. Pass it when configuring the MCP server.
Similar MCP servers
MCP servers catalog →engram by other authors comes first.
