Put Grok Bot in front of a Hermes agent so unattended jobs keep running

What you end up with

The reusable parts are the split (one system for conversation and approvals, another for unattended jobs) and the provenance footer, which lets you see at a glance what actually ran.

What you get

• A plain-language summary of the workflow
• A fill-in prompt template
• What to watch out for
• A link to the original source

Who it is for

Builders who already use AI agents and want a proven workflow they can copy into their own project.

Works with

Solid chips: the agent is named in this listing's own tools list. Dashed chip: Plain Markdown, so it also pastes into other agents such as Grok Bot, Hermes, OpenClaw or Muse. We list those as compatible by format only, not as tested.

Tools needed

Cost to run: Free with the agent plan you already use.
Time to result: One working session

How it works

1. List which jobs must run unattended and which need you in the loop

   List which jobs must run unattended and which need you in the loop

   Expected: Done before you move to the next step.

2. Run the unattended ones in Hermes with a separate profile per job

   Run the unattended ones in Hermes with a separate profile per job

   Expected: Done before you move to the next step.

3. Have a Grok Bot write and keep a small client for the Hermes local API

   Have a Grok Bot write and keep a small client for the Hermes local API

   Expected: Done before you move to the next step.

4. Pass every approval request back to Grok Bot instead of auto-approving

   Pass every approval request back to Grok Bot instead of auto-approving

   Expected: Done before you move to the next step.

5. Add a footer to each result with the model and tool calls used

   Add a footer to each result with the model and tool calls used

   Expected: Done before you move to the next step.

Sample prompt from the pack

Apply this workflow to [your project]:
1. List which jobs must run unattended and which need you in the loop
2. Run the unattended ones in Hermes with a separate profile per job
3. Have a Grok Bot write and keep a small client for the Hermes local API
4. Pass every approval request back to Grok Bot instead of auto-approving
5. Add a footer to each result with the model and tool calls used
Stop and ask me before anything irreversible: deploys, payments, sends or deletes.

Free preview of the file

# Put Grok Bot in front of a Hermes agent so unattended jobs keep running

PaperChase agent use case (free feed entry), version 1.0. Source: [Brad (@AFeastOfIdeas on X)](https://x.com/AFeastOfIdeas/status/2114713139564425276), 2026-10-26.

## What they did

The author split work between two systems. Hermes cron jobs run unattended, while scheduled Grok Bot routines that touch tools wait for approval, so long or nightly work moved to Hermes. A chief-of-staff Grok Bot wrote the script that talks to the Hermes local API: start a run, stream its events, poll until it finishes and pass approval requests back. Each Hermes instance has its own profile, a /new command starts a fresh context, and every reply carries a short proof-of-work footer naming the model and the tool calls used.

## Worth knowing

The reusable parts are the split (one system for conversation and approvals, another for unattended jobs) and the provenance footer, which lets you see at a glance what actually ran. Anything that runs unattended needs tight limits on what it can touch. Another builder posted a similar Grok Bot and Hermes setup the same week.

Worth knowing

• The reusable parts are the split (one system for conversation and approvals, another for unattended jobs) and the provenance footer, which lets you see at a glance what actually ran. Anything that runs unattended needs tight limits on what it can touch. Another builder posted a similar Grok Bot and Hermes setup the same week.
• Figures are as reported by the original author and were not re-measured by us.

Details

Version 1.0 · Updated Oct 27, 2026 · Markdown version of this page
License: Personal and commercial use. Credit the original source when you share the workflow. Download terms

Related

Prompt pack$14

AGENTS.md, CLAUDE.md and Cursor rules starter files

by PaperChaseWebb

Six fill-in instruction files for coding agents (base, Next.js plus Supabase, FastAPI, browser game, CLAUDE.md pointer, scoped Cursor rules) and 8 prompts to write, trim and test your own.

Works withClaude CodeCodexCursorPlain .md

Vibe coding · Beginner · 20 min setup

Prompt pack$19

Ask-first rules for AI agents: hard nos, approval cards and a test set

by PaperChaseWebb

Paste-in rules that make an agent ask before it sends, spends, posts or deletes, define what a real yes is, and stop retries after a no, plus 12 test scenarios to check it.

Works withClaude CodeCodexClaudeCursor+2Plain .md

Solo business · Beginner · 20 min setup