Skill Detail

Progressive Disclosure for Documentation

Restructure large documentation files (500+ lines, 5k+ tokens) into slim indexes with on-demand detail directories. Reduces always-included context by 40-60% while preserving full accessibility. Ideal for documentation auto-included in every agent context window.

Templates & WorkflowsClaude Code
Templates & Workflows Claude Code Published
โญ 14 GitHub stars
COPY SKILL INSTRUCTIONS (OPTIONAL)
npx skills add agentskillexchange/skills --skill progressive-disclosure Copy
Uses the third-party skills CLI, not an ASE-owned installer. Check your agent’s compatibility. This copies instructions; complete the upstream tool setup below separately.
At a glance
Author
gptme
Last updated
Jul 9, 2026
Quick brief

# Progressive Disclosure for Documentation

How it works

What this skill actually does

A pattern for restructuring large documentation files into slim indexes with on-demand
subdirectories, reducing always-included context by 40-60%.

## Problem

Large documentation files (500+ lines, 5k+ tokens) that are always included in context:
– Waste context budget on rarely-needed content
– Slow down response times
– Crowd out more relevant information
– Create cognitive overload

## Solution

Split monolithic docs into:
1. **Slim index** (~20% of original) – Always included, provides overview + navigation
2. **Detail directories** – On-demand loading when specific topics needed

## Structure

Before:
  TOOLS.md (~11k tokens always included)
  โ”œโ”€โ”€ Section 1 (rarely needed)
  โ”œโ”€โ”€ Section 2 (rarely needed)
  โ”œโ”€โ”€ ... (15+ sections)
  โ””โ”€โ”€ Section N (rarely needed)

After:
  tools/
  โ”œโ”€โ”€ README.md (~4k tokens always included)  # Slim index
  โ”œโ”€โ”€ topic1/README.md   # Detailed docs (~1k tokens each)
  โ”œโ”€โ”€ topic2/README.md   # Loaded on-demand
  โ””โ”€โ”€ .../README.md      # etc.

## When to Use

**Good candidates for progressive disclosure:**
– Files > 500 lines or 5k tokens
– Files always included in context (via config or auto-load)
– Files with distinct sections that are rarely needed together
– Documentation with both overview and detailed reference content

**Keep as single file when:**
– File is 10) | Group related sections into one subdir |
| Inconsistent depth (some sections slim, others still huge) | Apply pattern uniformly |
| Forgetting to update config/imports | Always audit auto-include lists after restructuring |
| Splitting files that are always read together | Only split genuinely independent sections |

## Installation

### Claude Code

Copy this skill directory:

cp -R skills/progressive-disclosure ~/.claude/skills/progressive-disclosure

Then invoke with `/progressive-disclosure` before restructuring a large documentation file.

### Manual

The pattern is pure documentation design โ€” no tools required. Read the Implementation
Guide above and apply the structure to any large file in your project.

## Related

– `autonomous-session-workflow` (in this catalog) โ€” auto-include optimization helps agents stay within context budget
– `five-element-spec` (in this catalog) โ€” scope documentation before restructuring