CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ponytail

Apply Ponytail's reuse-first ladder when the user invokes /ponytail, asks for the simplest implementation, or wants to avoid over-engineering and new dependencies in a coding task. Preserve all requested behavior and required checks. Not a default for every coding task, a prose-compression mode, or a replacement for the repository's debugging, testing, or refactoring workflow.

Quality

84%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Ponytail

Choose the smallest maintainable implementation that meets the complete request. This is Arnesto's scoped adaptation of DietrichGebert/ponytail; see provenance for the pinned source and differences.

Establish the behavior

Read the relevant code, callers, tests, and project conventions first. Identify the acceptance criteria, boundaries, and failure cases. For a bug, trace its root cause before choosing where to change it. Scope this skill to the current task; extend it across turns only when the user asks for a persistent mode.

Choose an implementation

Stop at the first option that satisfies the acceptance criteria:

  1. Omit speculative work outside the request.
  2. Reuse an existing helper, type, component, or established project pattern.
  3. Use the language's standard library.
  4. Use a native platform feature that meets the actual requirements.
  5. Use an already installed dependency.
  6. Write the smallest clear implementation needed.

Read the candidate's behavior before reusing it. A native date input, a database constraint, or a standard cache is suitable only if its accessibility, data, concurrency, lifetime, and product behavior match the request. A short snippet that misses those requirements is not a solution. Add a dependency when its benefit justifies its cost; explain that concrete tradeoff briefly.

Preserve quality and scope

  • Implement every requested acceptance criterion. Do not ship a reduced feature and ask the user to request the rest again.
  • Prefer readable code over one-liners. Avoid speculative abstraction, clever compression, unrelated deletion, and invented future extension points.
  • Keep validation at trust boundaries, security, accessibility, and error handling that protects data. Preserve documented compatibility constraints.
  • Follow the repository's test tools and required validation. For behavior changes, use its existing TDD workflow when practical; do not impose a one-test limit or replace its suite with ad hoc assertions.
  • Do not insert branded comments. Explain a non-obvious limitation only when the limitation matters to a maintainer.

Report the result

State what changed, why the chosen implementation fits, and what was verified. Name any remaining limitation. Match the user's requested explanation depth; Ponytail does not activate Caveman or change the communication language.

Savings are a hypothesis to measure, not a guarantee. Use the comparison cases in the evaluation plan before claiming improvement.

Repository

saski/arnesto

Path

.agents/skills/ponytail/SKILL.md

Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.