vikunja-mcp
Enables AI assistants to interact with Vikunja task management via ergonomic and raw REST tools for projects, tasks, labels, and comments.
README
@shichao402/vikunja-mcp
English | 简体中文
A Vikunja MCP server you can start directly with npx, using stdio transport for Claude Desktop, Cursor, Cherry Studio, and any other MCP-compatible client.
The current package exposes two layers:
- 19 ergonomic tools for day-to-day project, task, label, and comment workflows
- 161 raw REST tools generated from the bundled Vikunja Swagger snapshot, covering every documented operation in that snapshot
That means the package currently exposes 180 MCP tools in total. Raw tools use compact names shaped as vk_<method>_<resource>_<hash12>; each tool description still includes the exact REST operation so the model can call any Vikunja API endpoint precisely when needed.
Installation
No global install is required. The recommended distribution path is npx:
npx -y @shichao402/vikunja-mcp
Environment Variables
You must at least provide the Vikunja base URL:
VIKUNJA_BASE_URL=https://vikunja.example.com
API token authentication is the preferred option:
VIKUNJA_API_TOKEN=your_token
Supported environment variables:
VIKUNJA_BASE_URL=https://vikunja.example.com
VIKUNJA_API_TOKEN=your_api_token
# Optional alias
VIKUNJA_URL=https://vikunja.example.com
# Optional username/password login for self-hosted Vikunja
VIKUNJA_USERNAME=your_username
VIKUNJA_PASSWORD=your_password
VIKUNJA_TOTP_PASSCODE=123456
VIKUNJA_LONG_TOKEN=true
Notes:
VIKUNJA_API_TOKENworks for both Vikunja Cloud and self-hosted instances and has the highest priority.VIKUNJA_USERNAMEandVIKUNJA_PASSWORDare intended for self-hosted instances. The server will call/api/v1/login, cache the JWT, and reuse it.- If you pass a URL ending in
/api/v1, the server normalizes it automatically.
Claude Desktop Example
{
"mcpServers": {
"vikunja": {
"command": "npx",
"args": ["-y", "@shichao402/vikunja-mcp"],
"env": {
"VIKUNJA_BASE_URL": "https://vikunja.example.com",
"VIKUNJA_API_TOKEN": "your_api_token"
}
}
}
}
Tool Layers
Ergonomic tools:
vikunja_get_server_infovikunja_get_current_uservikunja_list_projectsvikunja_get_projectvikunja_create_projectvikunja_update_projectvikunja_delete_projectvikunja_list_tasksvikunja_get_taskvikunja_create_taskvikunja_update_taskvikunja_delete_taskvikunja_list_labelsvikunja_create_labelvikunja_list_task_labelsvikunja_add_label_to_taskvikunja_remove_label_from_taskvikunja_list_task_commentsvikunja_create_task_comment
Raw tools:
- One raw tool for each of the
161operations in the bundled Vikunja Swagger snapshot - Compact generated names, for example
vk_put_filters_dac0a4e19aa2,vk_post_tasks_2ca74cd58ce3,vk_put_projects_18c03cfa3b5a,vk_put_tasks_25d391326d4f - Path and query parameters are top-level fields
- JSON request payloads go in
body - Multipart uploads go in
form - File values inside
formuse{ "filename": "a.txt", "contentBase64": "...", "contentType": "text/plain" } - Binary download endpoints return
{ kind, contentType, filename, contentBase64, size }
API Coverage
Full coverage details and remaining limitations are tracked in English coverage docs and 中文覆盖文档.
Current status:
- Swagger operations in the bundled snapshot:
161 - Raw MCP coverage:
161 / 161 - Additional ergonomic tools:
19 POST /loginis exposed both as a raw tool and as an internal self-hosted login capability
Testing
Contract tests:
npm test
Live smoke tests against a real Vikunja instance:
VIKUNJA_LIVE_BASE_URL=http://127.0.0.1:34560 \
VIKUNJA_LIVE_USERNAME=mcpadmin \
VIKUNJA_LIVE_PASSWORD='StrongPass123!' \
npm run test:live
Local Development
npm install
npm run build
Show help:
node dist/index.js --help
Publishing
The package is intentionally distributed via npx because that is the lowest-friction path for MCP clients.
Publishing is automated through GitHub Actions, not local npm publish:
- Update
package.json,package-lock.json, andCHANGELOG.mdfor the new version. - Commit the release metadata, for example
chore: release v0.5.0. - Create and push a matching
v*tag, for examplev0.5.0. - The
Publish to npmworkflow publishes to npm via trusted publishing/OIDC.
The workflow verifies that the pushed tag version matches package.json before publishing.
GitHub repository: https://github.com/shichao402/vikunja-mcp
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.
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.
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.
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.