pinkbee-mcp
A read-only MCP server for Pinkbee that lets an AI assistant answer questions about rosters, shifts, and groups without making any changes.
README
pinkbee-mcp
Unofficial community integration; not affiliated with or supported by Pinkbee.
Proof of concept.
A read-only MCP server for Pinkbee
(*.mijnpinkbee.nl), an online platform for planning shifts and managing
volunteers.
It lets an AI assistant answer questions about a roster — which shifts still need people, who is scheduled, who is in which group — and nothing more: it cannot book, cancel or change anything.
Run it
docker compose up -d
That is all. It pulls withoutanickname/pinkbee-mcp and starts on built-in sample
data, so there is nothing to configure and no risk to anyone's real roster. The MCP
endpoint is then on http://127.0.0.1:8087/mcp.
Without compose:
docker run -d --name pinkbee-mcp -p 127.0.0.1:8087:8080 withoutanickname/pinkbee-mcp
To build from this checkout instead, uncomment build: . in docker-compose.yml and
run docker compose up -d --build.
Point your MCP client at that URL, or try it from the terminal:
curl -s -X POST http://127.0.0.1:8087/mcp -H "Content-Type: application/json" -H "Accept: application/json, text/event-stream" -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"pinkbee_check_connection","arguments":{}}}' | sed -n 's/^data: //p' | jq -r '.result.content[0].text'
The six tools
| Tool | What it answers |
|---|---|
pinkbee_check_connection |
Which data source is in use, and does it answer? |
pinkbee_list_groups_and_shifts |
Every group and shift with its id, and which groups may take which shift |
pinkbee_get_week_schedule |
The full roster for the week around each date you give |
pinkbee_list_open_shifts |
Shifts that still have open spots over the coming weeks |
pinkbee_list_group_emails |
Email addresses of the volunteers in a group (personal data) |
pinkbee_list_registrations |
Who signed up for which shift between two dates (personal data) |
Group and shift names are chosen by each Pinkbee administrator, so nothing here
assumes any particular naming: pinkbee_list_groups_and_shifts returns the whole
list and the assistant matches the names itself.
Connect it to a real Pinkbee
cp .env.example .env
Set these four, then docker compose up -d:
PINKBEE_DATA_SOURCE=live
PINKBEE_BASE_URL=https://your-organisation.mijnpinkbee.nl
PINKBEE_LOGIN=
PINKBEE_PASSWORD=
Use an account that only needs to read the roster. Every other setting is optional; see docs/deploying.md.
Safety
- Read-only. Every request to Pinkbee is a GET. The only POST in the whole codebase is the login itself.
- Mock data by default. Reaching a live instance needs a deliberate
PINKBEE_DATA_SOURCE=liveand credentials, or the server refuses to start. - Personal data is off by default. The two tools that return names or email
addresses refuse until you set
PINKBEE_ALLOW_PERSONAL_DATA=true. - A filter can never widen by accident. An id that does not exist is an error, an
unknown argument name is refused, and asking for every shift takes an explicit
all_shifts=true. Date ranges and id lists are capped. - Credentials only over HTTPS. A plain
http://Pinkbee URL is refused, since logging in sends the password. - Not open to the network. Optional bearer token,
Origin/Hostchecks against DNS rebinding, and the container publishes on127.0.0.1only. There is no TLS inside it, so put a reverse proxy in front if it has to leave the host. - One exception, on purpose:
PINKBEE_LOG_LEVEL=DEBUGalso turns on the libraries' debug output, which prints whole response bodies — so personal data does appear in debug logs. Use it while troubleshooting, not in normal running. See docs/deploying.md.
Documentation
| Document | Contents |
|---|---|
| docs/deploying.md | Every setting, TLS, health check, stdio mode, upgrading |
| docs/curl_examples.md | A worked curl example per tool, and two traps to avoid |
| docs/endpoints.md | The Pinkbee API: login, the endpoints used, how to find more |
| docs/login_sessions.md | How the session cookie is obtained, reused and renewed |
| docs/internals.md | Code layout, the mock data, running the tests |
Licence
MIT.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.