Stats
Actions
Tags
From claude-shorthand
Toggles LLMLingua-2 prompt compression on, off, dry-run, or shows current status and recent activity.
How this skill is triggered — by the user, by Claude, or both
Slash command
/claude-shorthand:shorthandThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Controls the LLMLingua-2 compression hook via `state.json`.
Controls the LLMLingua-2 compression hook via state.json.
STATE_DIR="$HOME/.claude/plugins/shorthand/bin"
LOG_FILE="$HOME/.claude/plugins/shorthand/compress.log"
mkdir -p "$STATE_DIR"
if [ "$ARGUMENTS" = "off" ]; then
echo '{"enabled": "off"}' > "$STATE_DIR/state.json"
echo "Shorthand compression disabled."
elif [ "$ARGUMENTS" = "dry-run" ]; then
echo '{"enabled": "dry-run"}' > "$STATE_DIR/state.json"
echo "Shorthand dry-run enabled — stats will be logged but prompts will not be compressed."
elif [ "$ARGUMENTS" = "status" ]; then
if [ -f "$STATE_DIR/state.json" ]; then
STATE=$(python3 -c "import json; print(json.load(open('$STATE_DIR/state.json')).get('enabled','on'))")
echo "Shorthand is: $STATE"
else
echo "Shorthand is: on (default)"
fi
if [ -f "$LOG_FILE" ]; then
echo ""
echo "Recent activity:"
tail -5 "$LOG_FILE"
fi
else
echo '{"enabled": "on"}' > "$STATE_DIR/state.json"
echo "Shorthand compression enabled."
fi
Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub gladehq/claude-shorthand