Skills toon
Compress JSON data to TOON format for ~40% context savings. Use when fetching APIs, reading JSON files, or any task outputting structured data. Pipe any output through `toon` - JSON gets compressed, non-JSON passes through unchanged.
install
source · Clone the upstream repo
git clone https://github.com/openclaw/skills
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/openclaw/skills "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/bonk-moltbot/toon" ~/.claude/skills/clawdbot-skills-toon && rm -rf "$T"
manifest:
skills/bonk-moltbot/toon/SKILL.mdsource content
TOON Context Compression
Pipe any command output through
toon to compress JSON and save ~40% tokens.
Usage
# API responses curl -s "https://api.example.com/data" | toon # JSON files cat data.json | toon # Any command - safe on non-JSON (passes through unchanged) some_command | toon
Install
# Copy script to PATH cp scripts/toon ~/.local/bin/ chmod +x ~/.local/bin/toon
Requires:
npx (Node.js)
Example
[{"id":1,"name":"Alice"},{"id":2,"name":"Bob"}]
→
[2]{id,name}: 1,Alice 2,Bob
When to Use
- Always when fetching JSON APIs
- Always when reading JSON files into context
- Safe to use on any output — non-JSON passes through
Reference
- Format spec: https://toonformat.dev
- CLI:
@toon-format/cli