DEV Community

Alexandre Lasly
Alexandre Lasly

Posted on

Running Hermes Agent on Android: A Production AI Agent in Your Pocket

Hermes Agent Challenge Submission

This is a submission for the Hermes Agent Challenge

No server. No cloud bill. Just a phone, Termux, and an AI agent that ships code.

When I first heard about Hermes Agent, the open-source agentic framework from Nous Research, I did what any reasonable developer would do: I tried installing it on my laptop. It worked. Then I asked myself — what if I could run it on my phone, always on, always ready?

Turns out you can. And it is surprisingly practical.

The Setup: Android + Termux

Termux gives you a functional Linux environment on Android. With it, you get bash, Python, Node.js, git — everything you need. Installing Hermes Agent is a one-liner:

pkg install python git
pip install hermes-agent
hermes setup
Enter fullscreen mode Exit fullscreen mode

The setup wizard walks you through picking a provider (I use DeepSeek), configuring tools, and connecting messaging platforms. Within 10 minutes, I had a fully operational AI agent running on a Samsung phone.

Why Run an Agent on Your Phone?

Three reasons:

  1. Always-on. Your phone is already on 24/7. No separate device, no VPS, no cloud bill.
  2. Messaging-native. Hermes Agent connects to Telegram natively — you interact with it like a contact. Wake up, check your messages, ask your agent what happened overnight.
  3. Capable. Despite running on ARM64 with limited RAM, Hermes performs real agentic work: multi-step reasoning, tool orchestration, file manipulation, git operations, web search, and cron scheduling.

How an AI Agent Let Me Work on an Entire GitHub Portfolio — From My Phone

This is where Hermes Agent truly shines. Over the past weeks, I built and deployed an entire professional portfolio site — portfolio design, HTML/CSS edits, Git versioning, GitHub Pages deployment, README documentation, and even a separate open-source project — all through my agent, without touching a laptop.

The agent did not just execute commands. It owned the workflow:

  • Repository creationgh repo create, initialize, push to main
  • Content edits — semantic find-and-replace via the patch tool, not fragile line numbers
  • Commits — meaningful commit messages in French and English, contextual to the change
  • Deployment — push to main → GitHub Pages auto-deploys. Live in seconds.
  • Cross-repo work — jumping between atlasnexusops.github.io (portfolio) and hermes-bounty-watcher (tool project) without losing context

The agent maintained persistent memory of the repo structure, file paths, and even my stylistic preferences ("use purple gradients, not green"). It remembered that I prefer French commit messages for the portfolio and English for open-source repos. That is something no CLI can do.

A real session: from idea to live site in minutes

Me: Remplace bot par IA agentique professionnelle dans le portfolio

Hermes: [searches for bot across the repo, finds line 223 in index.html, patches the exact string, commits with "Remplace bot par IA agentique professionnelle", pushes to main]

Hermes: Fait. C est en ligne.

Zero copy-paste. Zero SSH. Zero context switching. The agent had persistent knowledge of the repo structure, found the exact string, applied a semantic patch (not fragile line numbers), committed with a meaningful bilingual message, and pushed to main. GitHub Pages deployed within seconds.

Later that same session:

Me: Maintenant on supprime les coordonnees du footer et tu peux mettre mon nom en footer Alexandre Lasly

Hermes: [navigates to footer section, removes contact links, adds name, commits, pushes]

Hermes: Fait.

The whole GitHub surface area

Over multiple sessions, the agent managed:

Action Tool
Create repo + push terminal (gh CLI)
Edit HTML/CSS patch (semantic replace)
Write new files write_file
Search codebase search_files (ripgrep)
Read with context read_file (paginated)
Git commit + push terminal (git)
Schedule cron jobs cronjob
Deliver updates send_message (Telegram)
Remember preferences memory

Every GitHub operation — from init to deploy — went through the agent. No laptop. No IDE. Just Telegram messages on a phone.

Cron: The Secret Weapon

Hermes Agent is cron system deserves special attention. You define a job once, and it runs on schedule — fully autonomous:

hermes cron add "every 6h" \
  --script ~/.hermes/scripts/superteam_bounties.sh \
  --no-agent \
  --deliver origin
Enter fullscreen mode Exit fullscreen mode

The --no-agent flag means the script runs directly (zero LLM tokens) and its stdout is delivered verbatim. Silent when there is nothing to report.

Lessons Learned

  1. Memory is underrated. Hermes Agent is persistent memory means it remembers your preferences, project structure, and past fixes.
  2. Skills are living documentation. When you solve a complex workflow, save it as a skill.
  3. No-agent crons are efficient. For repetitive data collection, skip the LLM entirely.
  4. The phone is not a toy. An Android phone becomes a production-grade AI workstation.
  5. Agents own workflows, not just commands. The difference between a CLI wrapper and an agent is memory, planning, and context retention across sessions.

Try It Yourself

pkg install python git -y
pip install hermes-agent
hermes setup
Enter fullscreen mode Exit fullscreen mode

Connect a Telegram bot, schedule a cron, send a message — you will have an AI co-worker in your pocket within the hour.


Hermes Agent is open source. Runs anywhere you can install Python — including the phone in your hand right now.

Top comments (0)