Make your agent ask before it runs a risky command
In Hermes Agent the approval mode decides what happens to flagged commands. The default, smart, has a helper model approve low-risk ones, deny dangerous ones and send uncertain ones to you.
Source
@witcheer on X · 2026-10-16
Written in our own words. Credit to the original author. Link checked 2026-10-23.
What they did
In Hermes Agent the approval mode decides what happens to flagged commands. The default, smart, has a helper model approve low-risk ones, deny dangerous ones and send uncertain ones to you.
Result (as claimed)
You only get interrupted for the commands that genuinely need a human.
Tools they used
Cost: Hermes Agent: free plan, paid from $20/mo. Checked 2026-10-14.
Set it up
Map the task
Write the trigger, the inputs, the decisions and the output. Note which steps need a human yes.
Connect the tools
Set up Hermes Agent with the narrowest permissions that work. Use a test account or sandbox first.
Build the workflow
Create the steps in order. Add an approval step before anything sends, pays or deletes.
Run it on real but low-stakes data
Run a few times, read every output, and fix prompts or rules that misfire.
Schedule and monitor
Turn on the schedule or trigger, and add a daily summary or alert so you notice failures.
Copy the full setup prompt
Paste into your agent and it will walk you through the setup, asking before anything that sends, posts or spends.
You are helping me set up this AI use case end to end in Claude Code. Goal: Make your agent ask before it runs a risky command What the original author did: In Hermes Agent the approval mode decides what happens to flagged commands. The default, smart, has a helper model approve low-risk ones, deny dangerous ones and send uncertain ones to you. Reported result (as claimed by the author): You only get interrupted for the commands that genuinely need a human. Source: https://x.com/witcheer/status/2111052463058899099 Tools they used: Hermes Agent Category: Automation & business Please: 1. Ask me 3 short questions about my situation (accounts I have, budget, what I want the output to look like). 2. Propose a step-by-step plan using these tools or cheaper alternatives, with the time and cost of each step. Use official pricing pages; say "check pricing" if unsure. 3. Then do the setup with me one step at a time, waiting for my OK before anything that sends messages, posts publicly, spends money or deletes data. 4. Finish with a checklist to run it again next time.
You are helping me set up this AI use case end to end in Codex. Goal: Make your agent ask before it runs a risky command What the original author did: In Hermes Agent the approval mode decides what happens to flagged commands. The default, smart, has a helper model approve low-risk ones, deny dangerous ones and send uncertain ones to you. Reported result (as claimed by the author): You only get interrupted for the commands that genuinely need a human. Source: https://x.com/witcheer/status/2111052463058899099 Tools they used: Hermes Agent Category: Automation & business Please: 1. Ask me 3 short questions about my situation (accounts I have, budget, what I want the output to look like). 2. Propose a step-by-step plan using these tools or cheaper alternatives, with the time and cost of each step. Use official pricing pages; say "check pricing" if unsure. 3. Then do the setup with me one step at a time, waiting for my OK before anything that sends messages, posts publicly, spends money or deletes data. 4. Finish with a checklist to run it again next time.
You are helping me set up this AI use case end to end in Grok Bot. Goal: Make your agent ask before it runs a risky command What the original author did: In Hermes Agent the approval mode decides what happens to flagged commands. The default, smart, has a helper model approve low-risk ones, deny dangerous ones and send uncertain ones to you. Reported result (as claimed by the author): You only get interrupted for the commands that genuinely need a human. Source: https://x.com/witcheer/status/2111052463058899099 Tools they used: Hermes Agent Category: Automation & business Please: 1. Ask me 3 short questions about my situation (accounts I have, budget, what I want the output to look like). 2. Propose a step-by-step plan using these tools or cheaper alternatives, with the time and cost of each step. Use official pricing pages; say "check pricing" if unsure. 3. Then do the setup with me one step at a time, waiting for my OK before anything that sends messages, posts publicly, spends money or deletes data. 4. Finish with a checklist to run it again next time.
The Daily 5
Get five new use cases tomorrow morning
Free. Real sources. Unsubscribe any time.
Guides for this use case
How to share passwords and logins with an AI agent safely · How to make AI agents ask for approval before they act · How to set up Hermes Agent workflows · How to automate work with n8n and AI
Related use cases
Run a personal agent on Railway with Discord channels and an Obsidian memory
A founder runs Hermes on Railway, connected through his own MCP gateway to email, calendar, Slack, X, LinkedIn, GitHub, Obsidian and DNS, with code in a Vercel sandbox. He moved from Telegram to Discord for per-topic channels and has it keep a note on every person he works with in Obsidian.
Have an agent predict and book your weekly grocery restock from email receipts
He built an app in Codex that pulls receipts from his email into SQLite. Hermes uses that purchase history to predict what needs restocking, then logs into his supermarket every week, books a delivery slot and fills the cart. He can photo a barcode to add an item.
Control Hermes Agent on your Mac by voice from a wearable pendant
Bokko is a pendant prototype: hold a button, speak a task to Hermes Agent on your Mac, read the conversation on its 1.47-inch display and hear replies from the Mac. It uses Bluetooth, an 800 mAh battery and a 3D-printed case.
Use Hermes Agent as a daily Jarvis for coding, flights and sales
The author's Hermes Agent coordinates coding agents, works in their logged-in browser, finds flights and keeps a sales board current, with human approval at set points.