Let AI agents asynchronously ask users questions and receive text or voice replies.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "Duma" yet — see the docs or source repo.
When you find incomplete information while drafting product requirements, use Duma to ask me asynchronously for the target users, deadline, and priority; continue the PRD draft after I reply.
The AI first sends clarification questions, then produces a more complete PRD draft after receiving the user's text or voice replies.
Before running any script that may modify local files, use Duma to ask for my approval and list which files will change; continue only after I explicitly approve.
The AI sends an approval request first, waits for the user's reply, and then proceeds or stops based on the response.
If the analysis task runs longer than 10 minutes, use Duma to update me on progress and ask whether to continue deeper, export a summary, or switch to voice reply mode.
The AI proactively starts asynchronous check-ins during long tasks, letting the user provide next-step instructions by text or voice.
Store and retrieve development lessons locally with context-aware filtering for coding workflows.
Search messages, summarize context, and draft replies from chat history.
Run multi-agent debates and produce consensus recommendations with ranked options.
Enable MCP AI clients to see, act, and assist on macOS safely.
Orchestrate parallel AI agent workflows with dmux for faster development execution.
Create interactive videos and AI product demos that answer viewer questions.