Back to directory

Ponytail

An agent skill that makes AI coding agents write less code by checking, before writing anything, whether the code is needed, already exists, or is covered by the standard library or platform.

Ponytail

Ponytail is a skill for AI coding agents such as Claude Code and Cursor. Before writing code, the agent goes down a ladder of questions: whether the code needs to exist, whether it is already in the codebase, and whether the standard library, a native platform feature or an installed dependency covers it. Only then does it write the smallest version.

The skill keeps validation, error handling, security and accessibility code. The project publishes benchmarks from Claude Code sessions editing a FastAPI and React template, reporting less code, lower cost and shorter run time than the same agent without the skill, along with the method to reproduce them.

It is MIT licensed.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.