# Ponytail Gain

> benchmarks, ponytail, metrics, savings, scoreboard

- Skill: `lost-rob0t/ponytail-gain` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add lost-rob0t/ponytail-gain`
- Raw SKILL.md: https://api.skillmd.com/api/skills/lost-rob0t/ponytail-gain/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: lost-rob0t (https://skillmd.com/u/lost-rob0t)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/lost-rob0t/ponytail-gain

---


# Ponytail Gain

> External import of **Ponytail** by Dietrich Gebert and contributors, adapted from upstream v4.9.0 under MIT.

## Goal

Show Ponytail's published benchmark impact without pretending it measures the current repository.

## Output

Render this compact scoreboard:

```text
ponytail gain              upstream benchmark median · 5 tasks · 3 models

Lines of code   no-skill  100%
                ponytail    6–20%   down 80–94%
Cost            no-skill  100%
                ponytail   23–53%   down 47–77%
Speed           ponytail   3–6x faster
```

Then point repository-specific follow-up to `ponytail-debt` for counted deferrals and `ponytail-audit` for currently cuttable complexity.

## Rules

- These are upstream benchmark medians, not measurements of the current repository.
- Never invent a per-repo savings number: the unbuilt baseline does not exist.
- One-shot display only; do not change Ponytail mode or repository state.