> ## Documentation Index > Fetch the complete documentation index at: https://docs.usehenry.ai/llms.txt > Use this file to discover all available pages before exploring further. # Skills > Henry's memory is organized as skills: readable notes about your company, team, and workflows, with full history and one-click revert. Henry stores what it knows as **skills**: short, readable documents about your company, your team, each person, and the workflows you run. Henry reads the right skills at the right moment and updates them as it learns. The **Skills** page is rolling out gradually. If you do not see it in your sidebar yet, Henry is still building and using skills behind the scenes. Corrections work the same way: tell Henry what to change. ## The Skills page Open **Skills** to browse everything Henry knows about your workspace and where it came from. Skills are grouped as **Company**, **Team**, **Personal**, **Workflows**, and **Collections**. Each row shows when the skill was last used, when it was last verified, and its size. Badges mark skills that are **Always loaded** (Henry reads them in every conversation) and skills that are **Only you** (personal to you; other members never see them). ## A skill in detail Click a skill to read it. You see its description, the body, and, when present, a **Protected rules** callout: standing instructions Henry keeps word for word and always honors. Below the body is the **History**: every revision, who or what made it (a member, Henry, or the system), why, and when. Provenance labels include "from a member's statement", "from a tool result", "from shared content", "condensed by Henry", and "manual edit / undo". Expand a revision to see the exact diff. ## Revert Click **Revert** on any revision to restore what Henry knew at that point. Nothing is deleted. The revert is saved as a new revision, so you can undo it too. ## Editing Today the Skills page is read-only. To change something, tell Henry in chat and it updates the skill, with the change recorded in the history. ## How Henry keeps skills healthy * After conversations, Henry proposes small edits to the relevant skills and applies the ones that pass its checks: grounded in what was said, no secrets, no one else's private data, no conflicting rules. * When a skill grows too long, Henry condenses it, keeping protected rules intact and the full text in history. * Old time-sensitive facts are re-verified before Henry relies on them. Most of this is silent. On days when something needs a human, such as a fact Henry could not verify, workspace admins get one digest DM. Most days it sends nothing.