agentleFS
hotcode-devhotcode-dev/zerofactory/skills/zf-builder-ponytail/SKILL.md

zf-builder-ponytail

Skill2 starsChanged 40 days agoApache-2.0View on GitHub

Anti-overengineering code authoring playbook for zf-builder applying the 7-Rung Ladder of Laziness.

  • Installs packages

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 6 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Follow these steps on every task:

    line 26
  2. Distinctive:

    Understand Chesterton's Fence: Before modifying or deleting code, inspect why it exists (git log -S, comments).

    line 29
  3. Very specific:

    Search Before Building: Search the repository (search_files, grep) for existing utilities, constants, or helpers.

    line 30
  4. Very specific:

    Standard Library First (Rung 3): Always favor native language/stdlib modules (pathlib, json, subprocess, dataclasses, node:fs, crypto) over …

    line 33
  5. Very specific:

    Dependency Veto (Rung 5): Never run npm install <pkg> or pip install <pkg> unless the task explicitly requires an external integration that …

    line 34
  6. Very specific:

    Keep functions flat. Prefer early returns over deeply nested if/else blocks.

    line 36
  7. Distinctive:

    Avoid creating new classes, wrappers, or interfaces when a simple pure function does the job.

    line 37
  8. Distinctive:

    Do not write speculative configuration options or unused parameter flags.

    line 38
  9. Distinctive:

    Token Efficiency: Use targeted search/replace or concise patch hunks rather than rewriting whole files.

    line 42
  10. Distinctive:

    Verify with Tests: Run the targeted unit test suite to prove correctness before handing off.

    line 43

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

ZF Builder Ponytail — Code Authoring Playbook

"The best code is the code you never wrote."

This skill equips zf-builder with the Ponytail discipline: writing minimal, surgical, robust code that solves problems without bloat, unnecessary dependencies, or speculative abstractions.

The core decision framework is defined in the shared 7-Rung Ladder of Laziness.


The Builder's Authoring Workflow

Follow these steps on every task:

Step 1: Pre-Code Inspection (Rungs 1 & 2)

  1. Understand Chesterton's Fence: Before modifying or deleting code, inspect why it exists (git log -S, comments).
  2. Search Before Building: Search the repository (search_files, grep) for existing utilities, constants, or helpers. Never reinvent something that already exists in the project.

Step 2: Implementation Discipline (Rungs 3, 4, 5 & 6)

  1. Standard Library First (Rung 3): Always favor native language/stdlib modules (pathlib, json, subprocess, dataclasses, node:fs, crypto) over new libraries.
  2. Dependency Veto (Rung 5): Never run npm install <pkg> or pip install <pkg> unless the task explicitly requires an external integration that is impossible with existing libraries.
  3. Simplicity over Abstraction (Rung 6):
    • Keep functions flat. Prefer early returns over deeply nested if/else blocks.
    • Avoid creating new classes, wrappers, or interfaces when a simple pure function does the job.
    • Do not write speculative configuration options or unused parameter flags.

Step 3: Minimal Diff Hygiene (Rung 7)

  1. Surgical Edits: Touch only the exact lines required for the fix/feature. Never reformat, reorder, or touch unrelated lines.
  2. Token Efficiency: Use targeted search/replace or concise patch hunks rather than rewriting whole files.
  3. Verify with Tests: Run the targeted unit test suite to prove correctness before handing off.

Pre-Handoff Self-Check Checklist

Before calling hermes zerofactory move <task_id> done:

  • Did I add any code or parameters that aren't strictly required by the task? (If yes, delete them)
  • Did I add a new dependency? (If yes, can stdlib or an existing package do it?)
  • Is there an existing helper in the codebase I should have reused instead?
  • Are my diffs minimal, clean, and free of commented-out code or debug prints?

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.