agentleFS
saimskywalkersaimskywalker/lumenparty/skills/auron/SKILL.md

auron

Skill0 starsChanged 8 days agoMITView on GitHub

Load BEFORE writing or changing any code, by the main session or any builder. The code rules: read first, the laziness ladder (the least code that works), the never-cut list, the 300/500 line limits, write-it-right-once, fast by default.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Only then: the minimum that works, plus ONE small test when it has a branch, loop, parser, money or security.

    line 24
  2. Very specific:

    Follow the surrounding code (naming, comment density, layering).

    line 44
  3. Very specific:

    No N+1: one query or batched endpoint per list.

    line 51

≈ “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

Auron: no wasted motion

A legendary guardian wastes no swing. Write the least code that fully solves the task, shaped so it never needs a refactor. The ladder is adapted from the ponytail skill (DietrichGebert/ponytail, MIT).

1. Read first, then climb the ladder

Lazy about the solution, never about reading: open the code the change touches and trace the real flow. Then stop at the FIRST rung that holds:

  1. Does this need to exist at all?
  2. Is it already in this codebase? (search before writing a helper, widget, hook, provider, query, i18n key)
  3. Does the standard library do it?
  4. Is there a platform or framework feature? (a Material widget, a browser API, a router feature, the database itself)
  5. Is there a dependency already installed? Never add one for a few lines.
  6. Can it be one line?
  7. Only then: the minimum that works, plus ONE small test when it has a branch, loop, parser, money or security.

2. Never cut

Input validation at trust boundaries, data-loss handling, security, accessibility, error messages a user can act on, idempotency of money and counters. Small because necessary, never golfed. A deliberate shortcut gets // shortcut: <what> — revisit when <condition>.

3. File size (enforced by the Mini spell)

Soft 300 lines: stop adding concerns to the file; put the next cohesive piece beside it (a sibling widget or component, Dart part/part of, a module). Hard 500: a change may not take a file over it or grow one already over it. Generated files, i18n, migrations and tests are exempt. The fix is cohesion, never line golf.

4. Write it right once

Follow the surrounding code (naming, comment density, layering). switch over if/else ladders; one sealed / union state type over several booleans. Inline first, extract on the third real use. No drive-by refactors in a feature change (note them for cecil).

5. Fast by default (measured, not guessed)

No N+1: one query or batched endpoint per list. One poll per screen, opt-in. Flutter: const, select on providers, nothing heavy in build, dispose everything, images decoded at display size. Web: no layout shift, code-split routes, images sized and lazy, no waterfall fetches. Backend: index every new filter on a big table, paginate every list, never hold a lock across an await. Optimise a hot path only with a number.

6. Close every change

Name what was skipped, what was not checked, and any risk, in one or two lines.

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.