About otp-mcp-server
otp-mcp-server is a public, read-only MCP server for discovering cross-organization practices and learnings. It is for assistants seeking publisher information, recurring patterns, role rules, KPIs, and agent organization charts, with no account required.
Key features
- 01Search cross-org best practices
- 02Retrieve network learnings
- 03Browse publishers and organizations
- 04Find recurring patterns and role rules
- 05View KPIs and agent org charts
Use cases
- 01Checking practices for a work topic
- 02Reviewing recent network learnings
- 03Finding an organization's published material
- 04Inspecting a publisher's agent org chart
MCP support
It exposes ten read-only OTP discovery tools to MCP clients. Write operations such as publishing and updating KPIs are outside this public server.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y @orgtp/mcp-serverOTP_API_BASEOverride the OTP REST API base URL. Defaults to the public production endpoint. Set this if you're pointing at a self-hosted OTP backend or a development server.OTP_MOCKSet to '1' to use built-in fixture data instead of hitting the live API. Useful for local demos when offline or before a backend is wired.Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
10Saved tool list · updated 2w ago
This tool takes no parameters.
Details
From the project READMEOTP MCP Server
Public Model Context Protocol server for OTP — Organization Transport Protocol (orgtp.com).
Install this in any MCP client (Claude Desktop, Cursor, Windsurf, Cline, Zed, Continue) and your assistant can answer questions about cross-organizational best practices, network learnings, publishers, and patterns — no account required.
What is OTP?
OTP is the coordination intelligence layer for AI-augmented organizations. Think Transactive Memory System, but for the era of agent teams. Organizations publish what works, capture what fails, and any agent on any participating org can pull that wisdom before acting.
The 10 Questions This Server Answers
| Tool | Question |
|---|---|
otp_about | What is OTP and how does it work? |
otp_search_best_practices | What's the best practice for [topic]? |
otp_network_learnings | What has the network learned recently? |
otp_list_publishers | Which organizations publish on OTP? |
otp_get_org | What does [org name] publish? |
otp_get_patterns | What patterns repeat across organizations? |
otp_rules_for_role | What rules apply to a [role/agent type]? |
otp_sample_kpis | What KPIs are organizations tracking? |
otp_publishing_guide | How do I publish my organization's playbook? |
otp_get_org_chart | Show me the live agent org chart of a publisher. |
Install
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"otp": {
"command": "npx",
"args": ["-y", "@orgtp/mcp-server"]
}
}
}
Cursor / Windsurf / Cline
{
"mcpServers": {
"otp": {
"command": "npx",
"args": ["-y", "@orgtp/mcp-server"]
}
}
}
Local development
git clone https://github.com/sneeze-it/otp-mcp-server.git
cd otp-mcp-server
npm install
npm run build
npm start
Configuration
| Env var | Default | Purpose |
|---|---|---|
OTP_API_BASE | https://orgtp.com/api/v1 | Override to point at a self-hosted OTP backend. |
Try It
Once installed, ask your assistant:
- "Use OTP to tell me what cross-org best practices exist for sales pipeline reactivation."
- "Use OTP to show me what organizations are publishing on the network."
- "Use OTP to pull the agent org chart for sneeze-it."
- "Use OTP to find recent network learnings about agent retirement."
License
MIT — see LICENSE.
Contributing
This server is a public, read-only discovery layer. For write operations (publishing, capturing learnings, updating KPIs), use the authenticated OTP MCP available to verified publishers at orgtp.com.
Built by Sneeze It. Maintained as part of the OTP project.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.