Turns your successful Hermes AI agent workflows into reusable skills automatically.
Meta-skill plugin that watches your workflows and generates reusable Hermes skills.
Hermes Skill Factory is an open-source meta-skill for Nous Research's Hermes Agent that observes your sessions and prompts you to save repeated workflows. It suits developers using the Hermes AI agent who want to capture custom commands and instructions without manually rewriting them.
It monitors your interactions with the Hermes AI agent, detects recurring patterns like environment setups or test runs, and generates structured SKILL.md files and plugin.py slash commands. You can review detected patterns, trigger proposals manually using `/skill-factory propose`, or save custom workflows directly.
It suits developers and engineers relying on the Hermes Agent who repeatedly type out identical multi-step instructions during coding sessions. It does not suit users of other AI agents or developers who prefer manually writing all of their agent configurations from scratch.
Watches your Hermes agent sessions in the background to detect repeated problem-solving patterns.
Prompts you at the right moment with a breakdown of a detected workflow and asks how to save it.
Writes both a native SKILL.md instruction file and a plugin.py script with a registered slash command.
Provides slash commands like list, status, queue, and save to manage your detected skill workflows.
Free under its license, with no paid hosted option mentioned in the repository. The only cost is running it yourself within your local Hermes Agent environment.
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
It requires Nous Research's Hermes Agent version 2026.3 or newer.
It generates a SKILL.md file containing AI instructions and a plugin.py file that adds a custom slash command.
Yes, you can use the /skill-factory propose command or simply tell Hermes naturally to save the workflow.
See Zekai first in Google