Poppy
Local memory store that lets every coding agent recall your decisions and lessons
- Pricing
- Open source
- Platforms
- CLI
- Runs
- Locally
- Version
- 0.3.6
- Listing updated
- 3 days ago
About Poppy
Poppy is an open-source MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.
Memory lives in a small SQLite store under ~/.poppy on your machine, so no cloud, API key or per-query fee is involved. You or your agent save a decision, preference, fact or lesson in plain prose, and a recall query returns the few most relevant entries. Each record stores an id, type, content, project, source and capture date.
A single command per client wires it up. The site lists setup for Claude Code, Claude Desktop, Codex, Copilot CLI, Cursor, Gemini CLI, Goose and Hermes Agent, and the README adds VS Code, Windsurf and Pi. Some setups add lifecycle hooks and a primer block, so memory carries across sessions like /clear and /resume. With consent it can extract memories from sessions locally. Syncing to Trags is optional and off by default.
Install with pipx or uv (Python 3.11 or newer), check that poppy is on your PATH, restart the agent, then run poppy doctor to verify. It is released under Apache-2.0.
Install
Run in a terminal
claude mcp add poppy -- uvx poppy-memoryAdd to claude_desktop_config.json
{
"mcpServers": {
"poppy": {
"command": "uvx",
"args": [
"poppy-memory"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"poppy": {
"command": "uvx",
"args": [
"poppy-memory"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"poppy": {
"type": "stdio",
"command": "uvx",
"args": [
"poppy-memory"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"poppy": {
"command": "uvx",
"args": [
"poppy-memory"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"poppy": {
"command": "uvx",
"args": [
"poppy-memory"
]
}
}
}What it does
- Plain-prose memories
- Stores decisions, preferences, facts and lessons as short sentences.
- One store for many agents
- Claude Code, Cursor, Codex and others read the same local memory.
- One-command setup
- poppy setup <client> writes the MCP entry and, where supported, hooks.
- Automatic capture
- Optionally extracts durable memories from sessions locally, with consent.
- Diagnostics
- poppy doctor checks engine, storage, MCP config and hooks.
- Optional sync
- Pairing with Trags is separate and off by default.
Good for
- Keep a team's tooling preferences across agents
- Record a hard-won debugging lesson once
- Carry decisions across /clear and resumed sessions
- Share project context between Cursor and Claude Code
Pricing
Open source
Apache-2.0; local use needs no account
Prices change. Check trags.ai before you buy.
Poppy FAQ
Does it send data anywhere?
No by default. The store is local; cloud sync is an extra command.
What do I need to install it?
pipx or uv, with Python 3.11 or newer.
Which licence?
Apache-2.0.
Why do hooks fail with command not found?
The agent may not have poppy on its PATH; fix PATH and restart the agent.
Is Poppy free?
Yes. Poppy is open source, so you can use and self-host it for free.
How do I add Poppy to Claude Code?
Run this command in your terminal: claude mcp add poppy -- uvx poppy-memory
What are the best alternatives to Poppy?
The closest alternatives to Poppy are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.
Alternatives to Poppy

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

GitHub
MCP serverOfficial MCP server connecting AI assistants to GitHub repos, issues and workflows
Not checked yet- Remote
- Open source
- 34.5K GitHub stars
- Developer tools

Serena MCP: the IDE for your agent
MCP serverMCP toolkit giving coding agents symbol-level code retrieval and editing
Runs locally- Open source
- 31.5K GitHub stars
- Developer tools

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Ouroboros
MCP serverSpec-first layer that pins requirements and verifies results of AI coding agent work
Runs locally- Open source
- 6.4K GitHub stars
- Developer tools