Skip to main content
Glama
GoLinks

GoProfiles MCP Server

Official
by GoLinks
OverviewSchemaRelated ServersScoreDiscussions
Python
Hybrid

GoProfiles MCP Server

An MCP server that exposes GoProfiles as tools for AI assistants. Built with FastMCP.

Hosted HTTP Mode

This server is intended to run as a hosted remote MCP server over Streamable HTTP.

Public endpoint shape:

https://mcp.goprofiles.io/mcp

Local development endpoint shape:

http://localhost:8000/mcp

Related MCP server: KappaML MCP Server

Authentication

The hosted server does not use a shared GOPROFILES_API_TOKEN.

MCP clients should send a per-user GoProfiles OAuth/API bearer token with each request:

Authorization: Bearer YOUR_TOKEN

The MCP server forwards that header to the configured GoProfiles API (GOPROFILES_API_URL, default https://api.goprofiles.io). GoProfiles remains responsible for token validation, scope enforcement, refresh, storage, and revocation.

An OAuth client must be pre-registered in your GoProfiles workspace with:

  • Allowed scopes: profiles:read, profiles:write, search:read, users:read, bravos:read, bravos:write

  • Redirect URIs: the exact callback URL(s) your MCP client uses

Local Development

This project uses Python 3.12 and uv.

  1. Install Python 3.12.

  2. Install uv.

  3. Create the local environment and install dependencies:

uv sync

  1. Point the server at your GoProfiles sandbox API (not production). Copy the example env file and replace <your-username> with your dev path:

cp .env.example .env

See .env.example for every supported variable. .env is gitignored and must not be committed. Without a local .env, the server defaults to https://api.goprofiles.io.

  1. Run the MCP server locally:

uv run python -m goprofiles_mcp

Confirm startup prints your sandbox base, e.g. GoProfiles API base: https://dev.goprofiles.io/d/<you>/d/api.

The server binds to 0.0.0.0:8000 by default. Override with MCP_HOST and MCP_PORT if you need a different host or port:

MCP_HOST=127.0.0.1 MCP_PORT=9000 uv run python -m goprofiles_mcp

  1. Verify health and OAuth discovery:

curl http://localhost:8000/health
curl http://localhost:8000/.well-known/oauth-authorization-server

Test on ChatGPT

ChatGPT cannot reach localhost. Expose your local MCP server with a public HTTPS tunnel (ngrok), then add it as a custom connector.

  1. Start the MCP server locally (see Local Development) so it is listening on port 8000.

  2. Install ngrok (macOS example):

brew install ngrok

  1. Sign up at ngrok.com, copy your authtoken, and configure the agent:

ngrok config add-authtoken YOUR_AUTHTOKEN

  1. In a separate terminal, tunnel port 8000:

ngrok http 8000

  1. Copy the HTTPS forwarding URL ngrok prints (for example https://abc123.ngrok-free.app) and append /mcp:

https://abc123.ngrok-free.app/mcp

  1. Confirm the tunnel reaches your server:

curl -i https://abc123.ngrok-free.app/health

  1. In ChatGPT: Settings → Connectors, enable Developer Mode, then create a connector and paste the /mcp URL from step 5.

Notes

  • Keep both the MCP server and the ngrok process running while you test.

  • /mcp requires a Bearer token. Full ChatGPT OAuth login needs working GoProfiles authorize/token endpoints; without that, the connector may reach the URL but fail authentication.

  • A successful connect lists the tools in the Tools table below.

Docker

Build and run locally:

docker build -t goprofiles-mcp .
docker run --rm -p 8000:8000 goprofiles-mcp

Tools

Tool modules live under src/goprofiles_mcp/tools/ and are registered in src/goprofiles_mcp/server.py. Copy tools/_template.py when adding a new one.

Tool

Description

Scope

search_people

Search and filter the people directory by name, department, title, location, language, skill, interest, and group. Resolves a name to the uid other tools need.

search:read

get_profile

Full profile for one uid: bio, title, contact info, skills, interests, groups, languages, certifications.

profiles:read

search_celebrations

Recent and upcoming birthdays, work anniversaries, and new hires in one merged feed, filterable by type and day window.

profiles:read

get_availability

Whether one uid is available right now: local time, working hours, current meeting, open blocks left today, and next time off. Combines the workspace's connected calendar (Google or Outlook) with their configured working hours.

profiles:read

search_bravos

History of bravos already given, newest first, filterable by day window, person (giver or receiver), and giver/receiver department. Returns badge, when it was given, giver, receiver, points, and message.

bravos:read

search_bravo_types

The catalog of badge types that can be given, each with the bid preview_bravo needs.

bravos:read

preview_bravo

Resolves a recipient and badge and returns a preview for the user to approve. Stages the send but sends nothing.

bravos:read, profiles:read

create_bravo

Sends the Bravo the user approved from preview_bravo. Irreversible and notifies the recipient.

bravos:write, profiles:read

schedule_meeting

Create a staged calendar invite the user has explicitly approved. Organizer is the authenticated user. Irreversible — the event lands on both calendars and the invite is emailed.

profiles:write, profiles:read

This server cannot be installed

F
license - not found
Not graded
quality - not tested
B
maintenance

How are these scores calculated?

Maintenance

–Maintainers
–Response time
–Release cycle
–Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • garmin-mcp

    marcelohensantos
    F
    license
    Not graded
    quality
    B
    maintenance
    Exposes Garmin Connect data and workout management to AI agents, supporting tools, resources, and prompts for health data, workout creation, and coaching workflows.
    1

View all related MCP servers

Related MCP Connectors

  • GitHub

    Connect AI assistants to GitHub - manage repos, issues, PRs, and workflows through natural language.

  • mcp

    Connect your AI assistants to Keboola and expose your data, transformations, SQL queries, ...

  • Hint-Services-obsidian-github-mcp

    Connect AI assistants to your GitHub-hosted Obsidian vault to seamlessly access, search, and analy…

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/GoLinks/goprofiles-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server