Run a daily personal agent on a Raspberry Pi
The poster ran OpenClaw for about 13 weeks via Telegram on a Raspberry Pi, handling memory, crons, subagents, research workflows, APIs and scheduled tasks. ·
Done withOpenClaw
- What they did
- The poster ran OpenClaw on a Raspberry Pi for about 13 weeks as a Telegram-based personal assistant, with memory, cron jobs, subagents, research/summarization workflows, external APIs, reminders, and small automations across files and tools. Their pattern was that the human set direction, the main agent coordinated, subagents handled bounded analysis or coding, and the main agent verified; they also had to redesign brittle flows, like making child agents write results directly when parent readback failed.
- How it went
- It became daily infrastructure for scheduled checks, recurring analysis, debugging, config reviews, and follow-up, with the main win being continuity across days. It was still rough: model/config failures, update drift, shell quoting bugs, ACP runtime errors, and Pi resource pressure meant regular debugging and cleanup.
- Worth knowing
- The clearest catch was operational overhead: they archived one dead project after it bloated to about 1.6 GB, and they advise starting with one messaging channel, one model, one cron, and one real workflow.
Try it yourself with OpenClaw
Set up yourself as my daily assistant on my [Raspberry Pi or other always-on machine], reachable through [Telegram]. Keep persistent memory of my preferences, and schedule a daily morning message with [news topics, calendar, and tasks I care about]. Add one research workflow I can trigger by messaging you a topic, and confirm each scheduled task works by running it once. Before sending messages to anyone but me or spending money through any API, show me what you're about to do and wait for my approval.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: Reddit · May 18, 2026
Related use cases
Manage 18 agents from a Notion workspace
Notion used as the control plane for a fleet of 18 OpenClaw agents.
Automate home tasks like emails and ticket price checks
Microsoft's Omar Shahine used OpenClaw at home to draft emails and investigate concert-ticket prices.
Manage several websites across WordPress, Astro and Eleventy
The poster uses OpenClaw, alongside Claude Code and Codex, to manage and work across about five websites.
Get a second opinion on an MRI
One person's account of using Claude Code to analyze their own MRI.
Apply to 50 jobs in under 30 minutes
Claude computer use submitted 50 job applications in under 30 minutes, shown in a video demo.
Run boring parts of daily life with a local personal agent
The author runs a personal agent (Hermes) locally with local and cheap hosted models, saying it handles many boring parts of their life, though no specific tasks are given.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.