Skip to main content
affpapa.org
RU EN
skill Claude Code CX Codex Cursor GitHub Copilot

ponytail-gain — Claude Code skill: what it does, how to install

Displays measured benchmark savings across code volume, token spend, and execution time for Ponytail. Useful when evaluating the toolset.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓ 
npx skills add DietrichGebert/ponytail --skill ponytail-gain -a claude-code
Open on GitHub

What ponytail-gain does

The skill renders an ASCII scoreboard showing published benchmark data from Ponytail 5. The figures come from running headless Claude Code with Opus 5.5 across 39 tasks with 5 runs each, covering FastAPI and React features, bug fixes, and security tickets. Against the unassisted baseline, lines of code dropped by 53%, output tokens by 45%, cost by 26%, and time by 41%.

Execution is strictly one-shot. It writes no files, sets no flags, and never claims to calculate savings for your current repository. Because an alternative baseline does not exist in an active project, it avoids fabricated stats and points to /ponytail-debt instead.

Install this if you want quick reference data on Ponytail before applying it across projects. Skip it if you expect dynamic telemetry on your current workspace.

What it helps with

  • check Ponytail benchmark metrics for token and cost savings
  • review baseline comparisons across 39 sample agent tasks
  • verify test pass rates under Claude Code Opus 5.5 runs
  • find references to live debt and audit commands

Example prompts for the agent

  • /ponytail-gain
  • what does ponytail save
  • ponytail impact

How to install ponytail-gain

1Skills CLI (Vercel)

npx skills add DietrichGebert/ponytail --skill ponytail-gain -a claude-code

2Manually: put the files in place

~/.claude/skills/ponytail-gain/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/ponytail-gain/

From the same set6

The whole set →
ponytail DietrichGebert ponytail puts an agent into a senior-dev mode focused on the smallest complete change. Use it when you want clear replies and less code added to a project. Claude Code CX Codex Cursor +1 105,000installs ponytail-review DietrichGebert ponytail-review reviews a change beyond the diff, checking connected code, bugs, risks, load, tests, speed, and unnecessary code. Use it for PRs and commits. Claude Code CX Codex Cursor +1 56,300installs ponytail-audit DietrichGebert It audits a repository and ranks bugs, risks, and redundant code in a one-shot report. Use it when you need a review without code changes. Claude Code CX Codex Cursor +1 44,800installs ponytail-help DietrichGebert Shows a quick reference card for ponytail levels, skills, and configuration inside the chat. Useful when you need command triggers without leaving your editor. Claude Code CX Codex Cursor +1 42,300installs ponytail-debt DietrichGebert Turns every deliberate ponytail shortcut into one debt ledger: file, line, the ceiling the comment names, and the trigger for revisiting it. Claude Code CX Codex Cursor +1 42,200installs

Similar

obsidian-cli kepano A skill for working with Obsidian vaults from the command line: read and create notes, search, edit properties, debug plugins and themes. Claude Code CX Codex Cursor +1 80,800installs last30days mvanhorn Pulls posts and engagement on any topic from the last 30 days across Reddit, X, YouTube, TikTok, Hacker News, Polymarket, GitHub and the web. Useful when you want what people actually said, not what a blog summarized. Claude Code CX Codex Cursor +1 52,400installs marketing-psychology coreyhaines31 Applies behavioral science and mental models to marketing decisions. It helps teams analyze buyer motivations and solve funnel bottlenecks. Claude Code CX Codex Cursor +1 164,000installs content-strategy coreyhaines31 Plans content roadmaps, selects blog topics, and balances searchable search-intent pieces with shareable formats. Use it when deciding what to write next to bring in leads. Claude Code CX Codex Cursor +1 163,400installs find-skills vercel-labs Discovers and installs agent skills from skills.sh directly inside your coding assistant. Handy when you need domain-specific workflows for a project. Claude Code CX Codex Cursor +1 4,000,000installs grill-me Matt Pocock grill-me interviews you about a plan or design to expose gaps and sharpen decisions. Use it before implementation starts. Claude Code 1,424,476installs grill-with-docs Matt Pocock This skill interviews you about a plan or design and creates ADRs and a glossary along the way. Use it when decisions need sharper definition. Claude Code 1,219,756installs handoff Matt Pocock Creates a handoff document from the current conversation so another agent can continue the work. Use it when a task moves to a new session or agent. Claude Code 1,038,491installs

Questions about ponytail-gain

How do I install ponytail-gain in Claude Code?

Run npx skills add DietrichGebert/ponytail --skill ponytail-gain -a claude-code in your terminal. You can also save the file directly to ~/.claude/skills/ponytail-gain/SKILL.md.

What does ponytail-gain do?

It outputs an ASCII scoreboard summarizing benchmark savings in lines of code, tokens, and cost. It acts as a one-shot reference display without altering your code.

Does ponytail-gain measure savings in my current repo?

No, it displays fixed benchmark averages from 39 test cases. For repo-specific tracking of deferred shortcuts, run /ponytail-debt.

Source text

SKILL.md · MIT 
---
name: ponytail-gain
description: >
  Show ponytail's measured savings (code, cost, speed) from the benchmark.
  One-shot display. Use for /ponytail-gain, "what does ponytail save",
  "ponytail impact".
---

# Ponytail Gain

Display this scoreboard when invoked. One-shot: do NOT change mode, write flag
files, or persist anything.

The figures are the published agentic benchmark of Ponytail 5: headless Claude
Code (Opus 5.5, default effort) on 39 tasks (feature tickets in a real FastAPI +
React repo, bug fixes, security and privacy cases, small apps), 5 runs each,
against the same agent without the skill. 18 of the tasks have hidden checks
for correctness and safety. Each figure is the geometric mean of the per-task
medians. They are measured, not computed from the current repo.
Source: `benchmarks/results/2026-10-07-agentic.md` and the README.

## Scoreboard

Render plain ASCII bars. The bar length shows ponytail as a share of the
no-skill baseline; the label carries the exact figure:

```
  ponytail gain           benchmark · 39 tasks × 5 runs · Opus 5.5

  no-skill        ████████████████████  100%
  Lines of code   █████████···········   47%   ▼ 53%
  Output tokens   ███████████·········   55%   ▼ 45%
  Cost            ███████████████·····   74%   ▼ 26%
  Time            ████████████········   59%   ▼ 41%
  Hidden checks passed   97%  (no-skill 96%)
  Tests where the logic needs one   98%  (no-skill 68%)

  This repo:  /ponytail-debt  (shortcuts you deferred)
              /ponytail-audit (what's still cuttable)
```

## Honesty boundary

These are benchmark averages, not this repo. NEVER print a per-repo savings
number ("you saved X lines/tokens here"): the unbuilt version was never
written, so there is no real baseline to subtract from in a live repo. The
only real per-repo figures come from `/ponytail-debt` (a counted ledger), and
this card points there instead of inventing one.

## Boundaries

One-shot display. Edits nothing, changes no mode.

© DietrichGebert · MIT · DietrichGebert/ponytail

See also

Project sponsors

Companies that keep this analytics open
Gold sponsor
Nutra.Club

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor