Optimike Obsidian Tasks MCP
Enables AI-assisted task management by extracting and querying Obsidian Tasks from markdown files, with support for status mapping, date filters, and Obsidian Tasks query syntax.
README
Optimike Obsidian Tasks MCP
A Model Context Protocol (MCP) server for extracting and querying Obsidian Tasks from markdown files. Designed to work with MCP clients (Claude, Codex, IDEs, etc.) to enable AI‑assisted task management.
French version: README.fr.md
Prerequisites
- Node.js >= 16
- Obsidian Desktop
- Obsidian Tasks plugin configured in your vault: https://github.com/obsidian-tasks-group/obsidian-tasks
- Tasks config file present at:
<vault>/.obsidian/plugins/obsidian-tasks/data.json
Quickstart
npx optimike-obsidian-tasks-mcp /path/to/obsidian/vault
Features
- Extract tasks from Obsidian markdown files with compatibility for Obsidian Tasks settings
- Status mapping driven by your Tasks plugin config (core + custom statuses)
- Supports Tasks global filter + presets (optional)
- Date filters with relative ranges (EN/FR) and comparisons
- Optional Dataview task format parsing (
[due:: 2024-01-01], etc.) - Optional file metadata + frontmatter meta dates (with fallback to filesystem)
- Output as JSON or Markdown
Tasks plugin config (source of truth)
This server relies on the Obsidian Tasks plugin: https://github.com/obsidian-tasks-group/obsidian-tasks
This server reads your Tasks plugin settings (statuses, presets, global filter) from:
<vault>/.obsidian/plugins/obsidian-tasks/data.json
Make sure that file reflects your current Tasks configuration.
Tools
This MCP server provides the following tools:
list_all_tasks
Extracts all tasks from markdown files in a directory, recursively scanning through subfolders.
Input Parameters:
path(string, optional): Directory to scan. Defaults to the vault root.includePaths(string[], optional): Only include paths containing any of these substrings.excludePaths(string[], optional): Exclude paths containing any of these substrings.includeNonTasks(boolean, optional): Include NON_TASK statuses.includeFileMetadata(boolean, optional): Include file created/modified dates.includeMetaDates(boolean, optional): Include frontmatter meta dates (e.g., création/modification).metaFallbackToFile(boolean, optional, default true): If meta dates are missing, fall back to file dates.applyGlobalFilter(boolean, optional): Apply TasksglobalFilter.responseFormat("json" | "markdown", optional): Output format.responseLimit(number, optional): Limit number of tasks returned.useCache(boolean, optional, default true): Cache per-file parsing.
Returns: A JSON array of task objects, each containing:
{
"id": "string", // Unique identifier (filepath:linenumber)
"description": "string", // Full text description of the task
"status": "complete" | "incomplete" | "cancelled" | "in_progress" | "non_task",
"statusName": "string", // Optional - status name from Tasks config
"statusType": "string", // Optional - TODO | DONE | IN_PROGRESS | CANCELLED | NON_TASK
"filePath": "string", // Path to the file containing the task
"lineNumber": "number", // Line number in the file
"tags": ["string"], // Array of tags found in the task
"dueDate": "string", // Optional - YYYY-MM-DD format
"scheduledDate": "string", // Optional - YYYY-MM-DD format
"startDate": "string", // Optional - YYYY-MM-DD format
"createdDate": "string", // Optional - YYYY-MM-DD format
"doneDate": "string", // Optional - YYYY-MM-DD format
"cancelledDate": "string", // Optional - YYYY-MM-DD format
"metaCreatedDate": "string", // Optional - from frontmatter
"metaModifiedDate": "string", // Optional - from frontmatter
"fileCreatedDate": "string", // Optional - filesystem
"fileModifiedDate": "string", // Optional - filesystem
"priority": "string", // Optional - "high", "medium", or "low"
"recurrence": "string" // Optional - recurrence rule
}
query_tasks
Searches for tasks based on Obsidian Tasks query syntax. Applies multiple filters to find matching tasks.
Input Parameters:
path(string, optional): Directory to scan. Defaults to the vault root.query(string, required): Tasks query. Each line is treated as a filter.queryFilePath(string, optional): Used to resolve{{query.file.*}}placeholders.- All other parameters from
list_all_tasks(include/exclude, meta dates, global filter, etc.)
Returns:
A JSON array of task objects that match the query, with the same structure as list_all_tasks.
Supported Query Syntax (subset + extras):
-
Status filters:
done/not done(aligned with Tasks semantics)status.type is TODO|DONE|IN_PROGRESS|CANCELLED|NON_TASKstatus.name is "In Progress"
-
Date filters:
due|scheduled|start|created on|before|after YYYY-MM-DDdue|scheduled|start|created on or before|on or after <date>- Relative EN/FR:
today,tomorrow,yesterday,this week,next week,last week,aujourd'hui,demain,hier,cette semaine,semaine prochaine - Ranges:
due in next 7 days - Meta/file:
meta created before 2026-01-01,file modified after 2025-12-31
-
Tag filters:
no tags,has tagstag include #tag/tag do not include #tag
-
Path filters:
path includes string,path does not include stringfolder includes string,filename includes string
-
Description filters:
description includes string- Tasks with descriptions containing "string"description does not include string- Tasks with descriptions not containing "string"
-
Priority filters:
priority is high- Tasks with high prioritypriority is medium- Tasks with medium prioritypriority is low- Tasks with low prioritypriority is none- Tasks with no priority
Example Query:
not done
due before 2025-05-01
tag include #work
This would return all incomplete tasks due before May 1, 2025, that have the #work tag.
Usage
Installation
From npm (recommended):
# Install globally
npm install -g optimike-obsidian-tasks-mcp
# Or use directly with npx without installing
npx optimike-obsidian-tasks-mcp /path/to/obsidian/vault
From source:
git clone https://github.com/optimikelabs/optimike-obsidian-tasks-mcp.git
cd optimike-obsidian-tasks-mcp
npm install
npm run build
Running the Server
Using npm package (recommended):
# If installed globally
optimike-obsidian-tasks-mcp /path/to/obsidian/vault
# Or with npx (no installation required)
npx optimike-obsidian-tasks-mcp /path/to/obsidian/vault
From source:
node dist/index.js /path/to/obsidian/vault
You can specify multiple directories:
npx optimike-obsidian-tasks-mcp /path/to/obsidian/vault /another/directory
HTTP transport (optional)
The server also supports Streamable HTTP transport.
MCP_TRANSPORT_TYPE=http MCP_HTTP_HOST=127.0.0.1 MCP_HTTP_PORT=3011 \
npx optimike-obsidian-tasks-mcp /path/to/obsidian/vault
Session mode (default stateful):
MCP_HTTP_SESSION_MODE=stateful(auto sessionId)MCP_HTTP_SESSION_MODE=stateless(no session)
Performance & scope
To avoid scanning too much, use includePaths/excludePaths in MCP calls,
or set defaults via env vars:
MCP_TASKS_INCLUDE_PATHS(CSV): vault-relative paths, e.g.Efforts/Projets,Calendrier/NOTES PÉRIODIQUESMCP_TASKS_EXCLUDE_PATHS(CSV)MCP_TASKS_MAX_FILES(number)MCP_TASKS_CONCURRENCY(number, default 8)
Testing
To run the test suite:
npm test
See TESTING.md for detailed information about the test suite.
Inspection (MCP Inspector)
npm run inspect:stdio
# or
npm run inspect:http
Using with MCP clients (Claude, Codex, etc.)
Add this configuration to a client that supports MCP:
{
"mcpServers": {
"optimike-obsidian-tasks": {
"command": "npx",
"args": [
"optimike-obsidian-tasks-mcp",
"/path/to/obsidian/vault"
]
}
}
}
If you installed from source:
{
"mcpServers": {
"optimike-obsidian-tasks": {
"command": "node",
"args": [
"/path/to/optimike-obsidian-tasks-mcp/dist/index.js",
"/path/to/obsidian/vault"
]
}
}
}
Docker
Build the Docker image:
docker build -t optimike-obsidian-tasks-mcp .
Run with Docker:
docker run -i --rm --mount type=bind,src=/path/to/obsidian/vault,dst=/projects/vault optimike-obsidian-tasks-mcp /projects
Claude Desktop configuration:
{
"mcpServers": {
"optimike-obsidian-tasks": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"--mount", "type=bind,src=/path/to/obsidian/vault,dst=/projects/vault",
"optimike-obsidian-tasks-mcp",
"/projects"
]
}
}
}
Task Format
The server recognizes the following Obsidian Tasks format:
- Task syntax:
- [ ] Task description - Completed task:
- [x] Task description - Due date:
🗓️ YYYY-MM-DD📅 YYYY-MM-DD
- Scheduled date:
⏳ YYYY-MM-DD - Start date:
🛫 YYYY-MM-DD - Created date:
➕ YYYY-MM-DD - Priority:
⏫(high),🔼(medium),🔽(low) - Recurrence:
🔁 every day/week/month/etc. - Tags:
#tag1 #tag2
Example task: - [ ] Complete project report 🗓️ 2025-05-01 ⏳ 2025-04-25 #work #report ⏫
License
MIT License
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.