# Descriptions and Progressive Disclosure — Advanced Agent Workflows and Skills

Source: https://www.geekswithgeeks.com/en/agent-workflows/skills-descriptions

> Write descriptions that trigger correctly and keep detailed material in separate files.

## Load only what is needed

Only each skill's name and description sit in context all the time. The body loads when the skill is chosen, and extra files load only if the instructions ask for them. This **progressive disclosure** lets you install many skills without filling the context window.

## Weak vs strong description

A strong description says what the skill does and the situations that should trigger it, in the words a user would actually use.

```yaml
# Weak: too vague to trigger reliably
description: Helps with documents.

# Strong: says what and when
description: Convert a Markdown file to a styled PDF. Use when the user asks to export, print or share a .md file as PDF.
```

## Test the trigger

Try several phrasings of the request, including ones that should not match. If the skill fires too often or never, rewrite the description before touching the body.

**Quiz:** Why split long reference material into separate files?

- [ ] It is required by YAML
- [x] So it loads only when needed and saves context
- [ ] To make the skill run slower
- [ ] To hide it from the agent

*Answer:* So it loads only when needed and saves context. Large details loaded on demand keep the default context small.