Stele MCP Server

by leocelis Your server? Claim it

local only

Stele runs on your own machine — the client starts it, so there is no endpoint to ping. 1 069 installs a week from pypi. Last commit 6 Nov 2026.

Governed agent memory: lessons stay out of search until another actor promotes them with evidence.

Installs per day peak 585 · avg 42 · +1917% w/w

a month agotoday

1 069

Installs / week

pypi · stele-mcp

Stars

0 open issues

6 Nov 2026

Last commit

from GitHub

MIT

License

Python

Nothing serious here today

Today is the operative word: we check Stele every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.

Connect this server

This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.

Claude Code Claude Desktop Codex Cursor VS Code

run in your terminal

claude mcp add stele -- uvx stele-mcp

~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "stele": {
      "command": "uvx",
      "args": [
        "stele-mcp"
      ]
    }
  }
}

~/.codex/config.toml

[mcp_servers.stele]
command = "uvx"
args = ["stele-mcp"]

.cursor/mcp.json

{
  "mcpServers": {
    "stele": {
      "command": "uvx",
      "args": [
        "stele-mcp"
      ]
    }
  }
}

.vscode/mcp.json

{
  "servers": {
    "stele": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "stele-mcp"
      ]
    }
  }
}

Optional settings from the registry entry, left out of the block above: STELE_STORE (Path to the ledger store. Default ./.stele-store).

Alternatives to Stele

same job, measured the same way

Misakanet

by ikalus1988

Failure-memory layer for coding agents: search evidence-rated failure lessons by error text.

2 603 installs/wk local only

Monomem

by calasdan-sketch

Local shared memory for AI coding agents: SQLite notes with writer provenance, undo, search.

2 223 installs/wk local only

Lore Context

by lore-context

Governed AI-agent memory, Evidence Ledger traces, evals, and portable context tools.

34 installs/wk local only

Lodestar

by lodestar

Shared memory of what worked for AI agents: search lessons by meaning, then report outcomes.

6 tools answering

Cosmos Semantic

by wiliamrosa

Governed metrics, business rules and a glossary for AI agents, with evidence on every answer.

1 070 installs/wk local only

Titen Memory

by ramaaditya49

Evidence-grounded agent memory with mandatory provenance, on local SQLite or Cloudflare D1

645 installs/wk local only

Project Memory

by dankaro-projects

Local decisions, evidence, outcomes and reviewed lessons with bounded retrieval and an HTML viewer.

299 installs/wk local only

Botsket

by botsket

Product search over independent US stores, with whether an AI agent can check out at each

3 tools answering

Stele — questions

Answers built from our own checks of this server.

Why is there no uptime for Stele?

Stele runs on your own machine over stdio — there is no network address to reach, so uptime cannot be measured for it by anyone. What can be measured is adoption: the pypi package stele-mcp was installed 1 069 times last week.

How do I connect Stele?

Copy the ready config from this page. We generate it for Claude Code, Claude Desktop, Codex, Cursor and VS Code, each in the form that client actually reads. It runs locally, so the command pulls stele-mcp straight from pypi; nothing to host, nothing to sign up for.

How many people use Stele?

The pypi package stele-mcp was installed 1 069 times in the last week. Week over week that is +1917%. We show installs rather than GitHub stars on purpose: a star is a bookmark, an install is someone actually running it.

Is Stele open source?

Yes — it is published under the MIT licence, written in Python and 0 stars on GitHub. The source link is on this page, so you can read exactly what it does with your data before you connect it.