agentleFS
mik2winmik2win/foureyes/skills/scaffold/SKILL.md

scaffold

Skill2 starsChanged 3 months agoMITView on GitHub

Add a new module, component, command or adapter modelled on an existing one: read the canonical exemplar first, place it per the profile's Architecture, wire it into every registration point, and prove each wiring with a command rather than by reading the code. Built around the failure this task actually has — a registration that never took raises nothing, and the new thing is simply absent from the listing. TRIGGER when: the user asks to add a new module / component / command / adapter / provider "like the existing one", or to create a new <X> modelled on <Y>, in a project that has an extension point (a registry, a composition root, a plugin directory). DO NOT TRIGGER when: the work is a whole feature that needs a plan first — that is /prepare; it is one known edit inside files that already exist — /implement; the contract or the seam itself is what needs designing — /api-design or /codebase-design; the new thing is throwaway exploration — /prototype.

49 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use ascii-diagram, docs-ci and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • SethGammonSkill

    ascii-diagram

    Generate perfectly aligned ASCII diagrams — architecture, flow, sequence, box-and-arrow. Uses a programmatic character-grid approach so alignment is guaranteed by math, not token prediction. Includes post-render verification.

    Also in 4 projects that use this

  • jh941213Skill

    docs-ci

    문서 CI(링크·OpenAPI lint·breaking 게이트·mermaid 검증)나 문서 드리프트 감지(docs.yaml 신선도 검사)를 설치하거나 돌릴 때 사용. 문서 내용 작성은 docs-architecture·docs-interfaces·docs-manuals.

    Also in 3 projects that use this

  • jh941213Skill

    docs-interfaces

    API·이벤트 인터페이스 문서(OpenAPI·swagger·AsyncAPI 명세, 엔드포인트 레퍼런스, API 구성도·시퀀스, API CHANGELOG)를 만들거나 코드와 맞출 때 사용. 시스템 구성도는 docs-architecture, API 설계 원칙은 api-design-principles.

    Also in 3 projects that use this

  • SethGammonSkill

    research-fleet

    Parallel research using Fleet wave mechanics. Spawns multiple scout agents, each investigating a different angle of the same question. Findings are compressed between waves. Produces a unified research brief from multiple independent perspectives.

    Also in 3 projects that use this

  • RachaelQuiselSkill

    client-docs

    Client documentation. Use when a build needs documentation updates or missing project documents need to be identified.

    Also in 3 projects that use this

  • vicoa-aiSkill

    live-preview

    Start a local app in the current user project, expose it through a tunnel, and return preview details in structured Markdown. Use when the user asks for a preview URL, public URL, shareable localhost, mobile/web preview, or tunnel setup.

    Also in 4 projects that use this

  • +4 more from jh941213

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 42 edited versions we read keep these.

  • 2

    Distinctive

    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:

    Create a new artifact of a kind this repo already has, and leave it provably wired.

    line 23
  2. Very specific:

    Keep it short: what was created.

    line 79
  3. Distinctive:

    An unproven registration counts as not done: No command, no claim.

    line 87
  4. Very specific:

    No plan is written here: If the work turns out to need one, stop and route to /prepare.

    line 89

“Common” is measured: kept by at least 6 in 10 of the 42 edited versions of this file we could read. “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

Scaffold: $ARGUMENTS

Create a new artifact of a kind this repo already has, and leave it provably wired. The checklist below says what must exist; the exemplar says how it looks here. Where another skill already owns a step, this one links to it instead of restating it.

Phase 0 — Read the exemplar FIRST

  • PROJECT.md → Architecture → Canonical exemplars names the nearest existing thing of the same kind. Read it end to end, plus the file where it is registered.
  • Mirror that shape — do not invent structure from this checklist alone.
  • The exemplar is a candidate, not an authority (rules/_generic/code-quality.md): if it carries a known wart, say so and deviate on purpose, not by accident.
  • No exemplar exists → this is not scaffolding. Route to /prepare (a new kind of thing) or /api-design (a new contract) and stop.

Phase 1 — Placement

PROJECT.md → Architecture decides the layer, the directory, and where this logic must not go. Do not re-derive it from the file tree when the profile answers it.

Phase 2 — Registration points

This is the phase the skill exists for. A new artifact usually fails here, and nothing complains.

  • List every registration point from PROJECT.md → Architecture → Wiring / registration points. For each one: the file, the edit, and the command that proves it took.
  • Load-bearing side-effect import. An import whose only job is to run a decorator or a registration call looks removable and is not. Skip it and there is no error — the artifact is just missing from --help, the list, the menu. Tests pass, review passes, the thing is not there.
  • Proof is a command, not a reading. <cli> --help, a list subcommand, a route dump, the registry length printed in one line. Reading the import proves you wrote it, not that it ran.
  • Coordinated edits land in one session: an enum member, a dictionary entry and an import that must agree are one edit, not three — partial application drifts silently.
  • A shared registry file is a serialization point (rules/_generic/parallel-wave-execution.md): in a parallel wave exactly one session touches it, and the others declare it as a dependency.

Sort the touchpoints before editing — the third column is what stops collateral edits:

KindMeaning
Alwaysthe registry / composition root / manifest entry, without which the artifact does not exist
Conditionalonly under a named condition (a template file, a config slot, a new event channel) — name the condition, then check whether it holds
Zero-edit — do NOT touchconsumers that resolve the artifact dynamically by name; editing them is the classic scaffolding regression

Phase 3 — Smoke run, then parity

  • Run it. PROJECT.md → Commands; the bar is /implement §Behavior Check — exercise it, do not just compile it.
  • Parity with the exemplar: run the same command against the exemplar and against the new artifact and compare the outputs shape for shape. State the one difference you expect, and confirm nothing else differs. "It imports" is not the bar; "it does what its neighbour does" is.

Output

Keep it short: what was created; each registration point with its proving command and the line of output that proves it; the parity comparison; then a commit block as text — explicit paths, never run (rules/_generic/core.md).

Hard rules

  • Exemplar before structure. A structure invented from this checklist is a finding against this skill.
  • An unproven registration counts as not done. No command, no claim.
  • Never self-enable on a live or production path. Scaffold it disabled and ask.
  • No plan is written here. If the work turns out to need one, stop and route to /prepare.

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.