hermes-agent-spawning

Spawn independent Hermes Agent CLI processes for autonomous one-shot or interactive multi-turn tasks.

1|Updated Mar 12, 2026
One-click install
npx skills add https://github.com/kaminocorp/hermes-alpha-hunter --skill hermes-agent-spawning-kaminocorp
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hermes-agent-spawning
Source: https://github.com/kaminocorp/hermes-alpha-hunter/tree/main/skills/autonomous-ai-agents/hermes-agent
Command: npx skills add https://github.com/kaminocorp/hermes-alpha-hunter --skill hermes-agent-spawning-kaminocorp

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Running long autonomous tasks inside a single agent session blocks the parent conversation and limits tool access. This Skill launches fully independent Hermes Agent CLI processes with their own sessions, full tool access, and persistent logging, so long-running missions execute in the background without occupying the parent agent. ## Core Features & Use Cases - One-Shot Mode (-q): Run a single non-interactive query that executes autonomously and exits, ideal for fire-and-forget research or setup tasks. - Interactive PTY Mode: Launch a full interactive Hermes session for multi-turn collaboration, code review loops, and steered research with back-and-forth feedback. - Session Resumption: Resume previous CLI sessions with --continue or --resume to restore full conversation history from SQLite. - Use Case: Spawn two parallel agents — one building a FastAPI backend and another building a React frontend — then relay the API schema between them to keep both codebases in sync. ## Quick Start Spawn a background Hermes agent with the command hermes chat -q to research a topic and write the findings to a file, then monitor it with the process tool.

Frequently Asked Questions about hermes-agent-spawning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a Hermes agent task in the background?▼

Use one-shot mode with the terminal tool: run hermes chat -q followed by your task in quotes, with background set to true. The command returns a session_id you can monitor with the process tool while the agent works independently.

What is the difference between delegate_task and spawning a hermes process?▼

delegate_task spawns lightweight subagents sharing the parent process with a subset of tools and no session persistence. Spawning a hermes process launches a fully independent CLI with full tool access, its own SQLite session logging, and support for hours-long autonomous runs.

How do I resume a previous Hermes CLI session?▼

Run hermes --continue to resume the most recent CLI session, or hermes --resume followed by the session ID shown on exit. The full conversation history, including messages and tool calls, is restored from SQLite.

Why does submit not send my message in interactive PTY mode?▼

prompt_toolkit in raw mode expects a carriage return (\r) for Enter, but the submit action sends a line feed (\n), so text appears but never submits. The workaround is to run hermes inside tmux and use tmux send-keys with Enter, which sends the correct \r.

Can I run multiple Hermes agents in parallel?▼

Yes, spawn multiple background one-shot instances with separate hermes chat -q commands, each handling an unrelated task. Use the process tool's list action to track all running instances and check their output files when they complete.