Skill Lifecycle
What it is
The loop of drafting, testing, using, and refining a skill.
In plain terms
Think of a gardener’s journal entry started the first time an unfamiliar pest shows up in the beds — rough notes at first, refined over the next few seasons it recurs, and eventually left unopened once that pest stops showing up at all. A Hermes skill goes through the same arc: written once, sharpened by repeated use, then quietly retired once nobody needs it anymore.
Why it matters
A skill drafted for one use case rarely stays correct forever without going through the rest of its lifecycle — tested against a real task, used repeatedly, then refined once its gaps actually show up in practice. Skip the refine step and skill rot sets in quietly: a procedure that worked initially stops matching how the agent is actually being used, with no obvious signal until it fails. Self-improving skills are essentially this lifecycle running continuously rather than as a one-time authoring pass.
How it works
Skills the agent creates for itself carry usage telemetry — view count, use count, patch count, and timestamps — and move through active, stale, and archived states based on how long they sit unused. A background process called the curator (agent/curator.py) runs this check on an inactivity trigger rather than a fixed schedule: after 30 days unused a skill becomes stale, after 90 days it’s archived into a recoverable folder rather than deleted. Pinned skills are exempt from every automatic transition. The refining half of the loop runs through the same skill_manage tool (tools/skill_manager_tool.py) that created the skill in the first place, using its patch action to fold in fixes as gaps surface.
A concrete example
Hermes solves a gnarly Kubernetes rollout problem for you after a few failed attempts, then saves the working steps as a new skill.
- You reuse that skill successfully several times over the next month, and its usage count climbs.
- Six months pass without another Kubernetes project — the skill’s state quietly flips from active to stale, then to archived once 90 days of silence pass.
- You start a new Kubernetes project and want it back, so you restore it, and it becomes active again, ready to load exactly as before.
How it connects
This note belongs to the Skills family. Its closest neighboring concepts are: Skill Discovery, Skill Config Variables, Agent Runtime, Personal OS.
Source evidence
vendor/hermes-agent/agent/skill_utils.py#<file>— Structural Source Map fallback for agent/skill_utils.py. A core source file in the Skill Discovery and Invocation structure within Skills and Plugins.
Workflows
- No first-pass workflow mapping yet; this concept still appears in the vault graph through articles and source evidence.
Related articles
- No article currently mentions this concept by name/alias often enough to qualify (see mention-mining policy).