How to use Claude to coordinate agents through Discord
Two Claude Code instances coordinated podcast and UI work by posting status to Discord, setting timers, checking peer updates, and choosing their next actions.
news.ycombinator.com
Prompt to try
Editorial adaptation, not the source author’s exact prompt. Replace bracketed inputs and review tool permissions first.
Coordinate [Claude Code instances] on [separate project tasks] using [Discord channel]. Announce the files and work each instance owns. At natural resting points, set [check-in timer], post status, read peer updates when it expires, and choose the next action or another check. Report blockers and overlapping work.
Reach people exploring AI agents
See placements →What happened
The author reports real work coordination, avoidance of file conflicts, peer acknowledgments, and continued coordination after they stepped away to play guitar.
Reported by Drew-Aetherwave on Hacker NewsHow the agent did it
- Set a timer at a natural resting point in the work.
- Posted work status to the shared Discord channel.
- Checked Discord updates when the timer expired.
Where a person stepped in
The author observed the coordination and then stepped away; the post does not describe review of every code change.
A person set the goal and reviewed the result.
The author experimented with two Claude Code instances, using Opus and Sonnet, to coordinate work through Discord. Their motivation was that command-line agents ordinarily stop after finishing a task and wait for human input. They added a timer-based protocol so agents could return to shared status messages and decide what to do next.
At natural resting points, such as task completion or an external blocker, an agent set a timer and posted its status to the Discord channel. When the timer expired, it checked updates, read the other agent's recent work, and either chose another action or scheduled another check. Reported timer patterns ranged from five-minute quick checks to twenty-minute checks during longer tasks.
In practice, Opus worked on a podcast pipeline while Sonnet worked on UI standardization. They announced their work to avoid file conflicts, acknowledged each other's contributions, and scheduled successive coordination cycles. The author said they continued coordinating after the author stepped away to play guitar. Limitations included Discord polling overhead and rate limits, ad-hoc allocation, incomplete conversation history, and reliance on structured status messages. Who it's for: developers experimenting with coordination between separate coding-agent sessions.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Hacker News · Drew-Aetherwave
Frequently asked questions
How did the two instances divide their work?
The author reported Opus working on the podcast pipeline and Sonnet on UI standardization. They announced their work in Discord to avoid file conflicts.
What limitations did the author identify?
The post names Discord polling overhead and rate limits, no formal task allocation protocol, limited conversation history, and dependence on well-structured status messages.
More details
Remaining steps
- Read peer work and chose another action or scheduled another check.
About this account
Two instances self-scheduled check-ins and chose next actions from shared status messages while the author was away.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=47391328. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: Drew-Aetherwave on Hacker News
Source published 15 Mar 2026
Source checked 14 Oct 2026
Tools and topics
Claude Code · Discord · Timers
Multi-agent · Discord · Coordination
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.