Skip to content
MCP server

Worklog

Self-hosted MCP server where AI agents log finished work and you get searchable reports

Runs locally
Install

Worklog is an open-source MCP server for productivity. It runs locally on your machine and is started by your AI client.

Pricing
Open source
Platforms
CLI, Web, API
Runs
Locally
Version
1.1.1
Listing updated

Similar MCP servers

Install

Run in a terminal

claude mcp add worklog-mcp --env WORKLOG_OWNER='<WORKLOG_OWNER>' -- docker run -i --rm -e WORKLOG_OWNER ghcr.io/axute/worklog-mcp:1.1.1

Add to claude_desktop_config.json

{
  "mcpServers": {
    "worklog-mcp": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "WORKLOG_OWNER",
        "ghcr.io/axute/worklog-mcp:1.1.1"
      ],
      "env": {
        "WORKLOG_OWNER": "<WORKLOG_OWNER>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "worklog-mcp": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "WORKLOG_OWNER",
        "ghcr.io/axute/worklog-mcp:1.1.1"
      ],
      "env": {
        "WORKLOG_OWNER": "<WORKLOG_OWNER>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "worklog-mcp": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "WORKLOG_OWNER",
        "ghcr.io/axute/worklog-mcp:1.1.1"
      ],
      "env": {
        "WORKLOG_OWNER": "<WORKLOG_OWNER>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "worklog-mcp": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "WORKLOG_OWNER",
        "ghcr.io/axute/worklog-mcp:1.1.1"
      ],
      "env": {
        "WORKLOG_OWNER": "<WORKLOG_OWNER>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "worklog-mcp": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "WORKLOG_OWNER",
        "ghcr.io/axute/worklog-mcp:1.1.1"
      ],
      "env": {
        "WORKLOG_OWNER": "<WORKLOG_OWNER>"
      }
    }
  }
}

Using several servers? Build one config file with the MCP config generator.

Configuration

  • WORKLOG_OWNERYour user name; all entries belong to this userRequired
  • WORKLOG_AGENTAgent name stored with every entry, e.g. claude-desktop
  • TZTime zone for dates, e.g. Europe/Berlin

About Worklog

The README says worklog gives AI agents a place to record each finished result as its own entry, which you can later browse as a searchable log, generate weekly or monthly reports from, or turn into a timesheet. Entries are stored in SQLite inside a single container, and the server speaks Streamable HTTP or stdio to MCP clients such as Claude Code, Claude Desktop and claude.ai.

A web view adds full-text search, filters, an activity calendar, timesheet rounding, project notes and duplicate detection. Each agent gets its own token, and the agent name is taken from that token, so entries can be traced back to the client that wrote them. Several users can share one instance while their data stays separate.

This suits developers and knowledge workers who use AI coding or chat assistants daily and want a record of what those sessions produced, along with a way to export it as Markdown, CSV or JSON.

What it does

Agent work logging
MCP tools let agents log, correct and search entries, one per finished result.
Web view and search
A browser interface offers full-text search, filters, an overview calendar and a timesheet with rounding.
Reports and notifications
Weekly or monthly reports can be saved as files or sent to services such as ntfy, Discord, Telegram and e-mail via Apprise.
Import and export
Data can be exported as Markdown, CSV or JSON, and imports cover JSON from other instances and earlier Claude Code, claude.ai and JetBrains Junie sessions.
Per-agent tokens
Each agent uses its own token stored as a hash, and several users are kept strictly separated.
Small footprint
Runs as one container on SQLite with images for amd64 and arm64.

Good for

  • Keeping a running record of what an AI coding assistant finished each day
  • Generating weekly or monthly work reports from logged entries
  • Producing a timesheet with rounding from agent activity
  • Backfilling a log for an old project from its git history

Pricing

Open source

MIT licensed, self-hosted.

Worklog FAQ

Which environment variables does the MCP server require?

The registry entry lists WORKLOG_OWNER as required. WORKLOG_AGENT defaults to stdio and TZ defaults to Europe/Berlin.

Which transports does Worklog support?

The registry entry lists stdio, and the README describes HTTP (Streamable HTTP) as well.

Where is the data stored?

The README says it uses SQLite in one container, with the database path set through configuration.

Is Worklog free?

Yes. Worklog is open source, so you can use and self-host it for free.

How do I add Worklog to Claude Code?

Run this command in your terminal: claude mcp add worklog-mcp --env WORKLOG_OWNER='<WORKLOG_OWNER>' -- docker run -i --rm -e WORKLOG_OWNER ghcr.io/axute/worklog-mcp:1.1.1

What are the best alternatives to Worklog?

The closest alternatives to Worklog are Todo MCP, ShiftLog.it, Monologue, AgentRQ and Slates.

Where this comes from

  • Description: written from the project's README and its MCP Registry entry.
  • Install steps: from the official MCP Registry entry.

How AIPoweredBot checks servers