Erkle team memory
Shared, reviewed memory that Claude Code, Cursor and Codex agents search before a task
- Pricing
- Free trial
- Platforms
- API, CLI, self-hosted
- Runs
- Remotely
- Version
- 0.1.1
- Listing updated
- 3 days ago
About Erkle team memory
Erkle team memory is an MCP server for knowledge and memory. It runs as a hosted remote server, so there is nothing to install locally.
Erkle stores a team's decisions, conventions, constraints and lessons in a hosted memory that every teammate's agent can query. An agent searches before a task and proposes what it learned afterwards; a person approves proposals from a review queue before they go live, and every memory shows its author, the agent that wrote it and a source link.
Memories are scoped team, project or repo, and a search from a repo reads its own scope, its projects and the team's. Memories you cannot read do not appear at all, not even as access denied. Every edit is a new version, conflicting edits are rejected, writes containing private keys or tokens are refused, and the usage log never stores search text or memory content.
npx @erkle/cli init writes the MCP entry for Claude Code, Cursor and Codex and adds an instruction block to AGENTS.md or CLAUDE.md. For sandboxes and CI, an orchestrator can mint short-lived read-only keys per agent. Memories stay on Erkle's servers, and no AI provider receives them for embedding.
Install
Run in a terminal
claude mcp add --transport http erkle https://erkle.ai/mcpAdd to Settings → Connectors → Add custom connector
https://erkle.ai/mcpAdd to ~/.cursor/mcp.json
{
"mcpServers": {
"erkle": {
"url": "https://erkle.ai/mcp"
}
}
}Add to .vscode/mcp.json
{
"servers": {
"erkle": {
"type": "http",
"url": "https://erkle.ai/mcp"
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"erkle": {
"serverUrl": "https://erkle.ai/mcp"
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"erkle": {
"type": "http",
"url": "https://erkle.ai/mcp"
}
}
}Tools this server provides (2)
Taken from the project documentation; the running server may differ.
- memory_search
- Searches team memories scoped to a repo, project and team.
- memory_create
- Proposes a new memory for maintainer review.
What it does
- Review before reuse
- Agent writes wait for a person's approval by default.
- Scoped memory
- Team, project and repo scopes that never reveal hidden memories.
- Version history
- Each edit is a new version; conflicts are rejected.
- Secret guard
- Rejects writes that contain private keys or tokens.
- Per-agent keys
- Orchestrators mint limited, expiring keys and every call is logged by key.
- One-command setup
- `erkle init` configures Claude Code, Cursor and Codex.
Good for
- Stop one agent repeating a mistake another already learned
- Record team conventions that CLAUDE.md files cannot share across repos
- Give CI or sandbox agents scoped read-only access
- See which memories agents actually use
Pricing
Free trial, from $5 a month
$5 per user per month or $50 per user per year; 7-day trial, no card.
Prices change. Check erkle.ai before you buy.
Erkle team memory FAQ
What does it cost?
$5 per user per month, or $50 per user per year, with every feature included and no metering; the trial is 7 days.
Can we self-host it?
Yes, priced separately; it runs as two small services plus Postgres.
Do agents write directly?
They propose; a person reviews by default.
Is there an API key to copy?
Agents sign in through the browser; keys are for headless setups.
Is Erkle team memory free?
Erkle team memory is a paid product with a free trial. Paid plans start at $5 a month.
How do I add Erkle team memory to Claude Code?
Run this command in your terminal: claude mcp add --transport http erkle https://erkle.ai/mcp
What are the best alternatives to Erkle team memory?
The closest alternatives to Erkle team memory are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.
Alternatives to Erkle team memory

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