Who it is forA person curious about Taskuary who wants to see the workflow before connecting work mail, chats or tickets.
The problemSetting up an AI assistant that touches real mail is a lot to commit to before you know how the task flow feels.
The outcomeYou have seen one fictional request move through triage, agent work and your approval, with no outside system touched.
What you need first
- Python 3.10 or newer, or the Windows Taskuary.exe download, or Docker
- An AI provider key or a local Ollama model
- At least one inbound channel such as mail, chat or tickets
- A coding CLI such as Claude Code, Codex or Gemini if you want agents to do the work
Steps
- Install Python 3.10 or newer on Windows, macOS or Linux.
- Run
pip install taskuary - Run
taskuary --demo - Open http://127.0.0.1:7787 and choose Walk me through my tasks.
- Follow Ruth's vendor spend request to the Approve & send step and stop there.
Prompt for your agent
In the ldbumble/taskuary repo, install the package with `pip install taskuary` and start it with `taskuary --demo`. Report the title of the first task the Assistant shows, and confirm that the demo cannot send messages or start agents. Verify the demo output by checking that the task names fictional data only.
Caveat. The demo uses fictional data and scripted replies, so it does not show how your own mail or tickets will be triaged.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for ldbumble/taskuary and "Try the guided demo before connecting anything"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer