> For the complete documentation index, see [llms.txt](https://docs.rox.com/development/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.rox.com/development/product/chat/chat-skills.md).
# Chat Skills
## Skills
Skills are reusable instructions you save once and reuse in Chat. A Skill captures how you want Chat to handle a specific kind of request — a research format, an outreach style, a way of summarizing calls — so you don't have to re-explain it every time.
Each Skill has three parts:
* **Title** — a short name for the Skill.
* **Trigger** — a description of *when* Chat should use the Skill.
* **Instructions** — what Chat should do when the Skill is used.
### How Skills are used
A Skill is applied to Chat in two ways:
* **Automatically** — when a Skill is enabled, Chat uses it on its own whenever your request matches the Skill's **Trigger**.
* **On demand** — type `/` in the chat composer to open the Skill picker, then choose a Skill to apply it to that message. You can add your own text around it.
### Create a Skill
1. In the chat composer, click the **+** icon and select **Skills**.
2. Click **New skill**.
3. Fill in the form:
* **Title** — name the Skill (for example, "Account one-pager").
* **Trigger** — describe when Chat should use it (for example, "When I ask for a summary of an account").
* **Instructions** — write what Chat should do. Use the formatting toolbar for headings, lists, and links. The clearer the instructions, the more consistent the results.
* **Visibility** — choose **Private** (only you) or **Everyone in your org**.
4. Click **Create skill**.
{% hint style="info" %}
Write the **Trigger** the way you'd actually describe the task in Chat. A vague trigger makes it harder for Chat to know when to apply the Skill automatically.
{% endhint %}
#### Attach files
You can attach supporting files to a Skill — for example a template, a style guide, or reference data — and Chat will use them when the Skill runs.
* **Supported types:** PNG, JPG, GIF, WebP, PPTX, TXT, MD, Markdown, CSV, TSV, JSON, JSONL, YAML, YML
* **Max size:** 50 MB per file
Drag files into the **Bundle Files** area, or click to browse. You can add or remove files before saving.
### Enable and disable Skills
Each Skill has a toggle. When a Skill is **disabled**, Chat will not use it and it cannot be invoked through a `/` command.
Skills are grouped into two sections:
* **Personal** — Skills you created.
* **Team** — Skills shared by teammates across your org.
You control your own toggles independently — turning a Team skill off only affects you.
### Share a Skill with your org
Set a Skill's **Visibility** to **Everyone in your org** to make it available to everyone. Shared Skills appear in each teammate's **Team** section, where they can enable them for themselves. Only the Skill's owner can edit or delete it.
### Admins: enable a Skill by default
When creating or editing an org-wide Skill, admins can turn on **Enable by default for all members in my org**. This switches the Skill on for everyone automatically. Members can still turn it off for themselves.
### Tips
* Keep each Skill focused on one task — smaller, specific Skills are more reliable than one that tries to do everything.
* Make the **Trigger** match how you naturally phrase the request.
* Spell out the desired **format** in the instructions (length, structure, tone) so results stay consistent.
### Examples
Two of the most popular uses for Skills are enforcing a **writing style** and applying **brand guidelines** — anything you'd otherwise have to re-explain or correct on every request. A Skill turns those standards into rules Chat follows automatically.
#### Example: Match my writing style
Use this when you want everything Chat drafts to sound like you.
* **Title:** My writing voice
* **Trigger:** When I ask to draft or rewrite an email, message, or note
* **Instructions:**
> Write in my voice: direct, warm, and concise. Use short sentences and active voice. Skip filler openers like "I hope this finds you well." Prefer plain language over jargon, and never use exclamation points. Open with the point, then the context. Sign off with just my first name.
#### Example: Apply our brand guidelines
Use this when deliverables need to follow your company's brand standards. State the rules as hard requirements so they're applied consistently.
* **Title:** On-brand decks and docs
* **Trigger:** When I create or edit a slide deck or document
* **Instructions:**
> Apply these brand rules as hard requirements to everything you produce:
>
> * **Voice:** confident, clear, and professional. No all-caps text anywhere — use sentence case for titles and headings.
> * **Colors:** use our brand palette — primary `#0B5ED7`, accent `#BDF0FF`, dark `#00126E`. Don't introduce off-brand colors.
> * **Logo & footer:** place the logo bottom-right on content slides; keep a small page number in the footer.
> * **Cover & closing:** open with a branded cover slide and end with a clean closing slide showing the logo centered.
>
> If a rule can't be applied exactly, get as close as possible rather than dropping it.
{% hint style="info" %}
The more specific your instructions — exact colors, tone rules, structure — the more consistent the results. Attach a brand asset or template as a file and Chat will use it when the Skill runs.
{% endhint %}
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.rox.com/development/product/chat/chat-skills.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.