Skip to content

DoneThat

Verified

Privacy-first work tracking with AI summaries and long-term memory.

MCP ServerAI & KnowledgeRemote (streamable-http)
View source
Updated 2026-06-16

What is the DoneThat MCP server?

DoneThat enables users to log work without compromising personal data. It processes activity into structured summaries and reports that can be fed into AI models for coaching or analysis.

Built for privacy, the server keeps memory local or encrypted and exposes only necessary context to connected AI clients through the MCP protocol.

Install & connect

Add this to your MCP client config. Pick your client below and copy.

{
  "mcpServers": {
    "donethat": {
      "url": "https://mcp.donethat.ai/"
    }
  }
}

Example prompts

Once connected, try asking your AI client:

Summarize my work from the past week and highlight focus areas.
Generate a coaching report based on my recent task patterns.
Export my long-term memory as structured data for analysis.
Create a privacy-safe monthly productivity report.

Security & permissions

Requires streamable-http connection to a remote DoneThat instance; may need API keys or tokens configured via environment variables to access stored work data.

What you can do with DoneThat

Daily Work Summaries

Automatically compile end-of-day reports from tracked tasks and time entries.

AI Coaching Sessions

Feed long-term memory into an AI assistant for productivity advice and goal tracking.

Privacy-Preserving Reports

Generate shareable reports without exposing raw activity data to third parties.

How to use DoneThat

  1. 1Install the DoneThat MCP client package in your environment.
  2. 2Configure the streamable-http endpoint URL and any required credentials.
  3. 3Add DoneThat to your AI client's MCP server list.
  4. 4Grant the server access to your local work tracking data source.
  5. 5Test connectivity by requesting a sample summary from your AI.

DoneThat: pros & cons

Pros

  • +Strong emphasis on user privacy and data control.
  • +Provides structured memory suitable for long-term AI use.
  • +Generates actionable reports and coaching without manual effort.
  • +Works over standard streamable-http for easy remote access.

Cons

  • No tools are documented in available descriptions.
  • Requires separate work-tracking data source to be useful.
  • Remote transport may introduce latency compared to local servers.
Did you find this helpful?

Frequently asked questions

DoneThat prioritizes privacy and can keep memory on your own infrastructure.

User reviews

Verified reviews from the community shape this listing's rating.

Loading reviews…

Sign in to review

Promote DoneThat

Add this badge to your website, or share the tool.

DFeatured on DhanasviDoneThat 0