agentleFS
sayurionishisayurionishi/playwright-scaffold/.claude/skills/ai-native-workflow/SKILL.md

ai-native-workflow

Skill1 starsChanged 3 months agoMITView on GitHub

Use to understand the overall way of working in this scaffold — the conversation contract, the task lifecycle, and the principles that keep AI-generated output consistent. The operating manual that ties the other skills together.

1 other project uses a version of this.

Adopted by

Most recently taken up by

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. Very specific:

    No code before approval (except trivial Direct-Mode fixes, premise verified first).

    line 20
  2. Distinctive:

    ASK. This is the single mechanism that prevents invented selectors, endpoints, and folders.

    line 26
  3. Very specific:

    If PROJECT.md is missing, run bootstrap.

    line 43

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

AI-Native Workflow (operating manual)

This skill is the why behind the others. CLAUDE.md is the table of contents; the per-task skills are the deep rules; this explains how they fit.

Three layers

  1. Constitution (CLAUDE.md §1) — hard stops. Refusals, not preferences.
  2. Specialized skills (.claude/skills/*) — deep, phased rules, loaded by the wording of the task.
  3. Code conventions — the actual TypeScript, read on demand. The code is the truth.

The conversation contract — audit-then-edit

State → route+explore → propose (with confidence 1–10) → human approves → apply → verify → report. No code before approval (except trivial Direct-Mode fixes, premise verified first). See the Universal Spine in AI-WORKFLOWS.md.

The confidence gate

Phase-4 proposals carry a 1–10 confidence. < 5 means you're guessing — don't propose, go explore or ASK. This is the single mechanism that prevents invented selectors, endpoints, and folders.

Five principles that make output consistent

  1. Single source of truth per value class — URLs/secrets in config (env); paths/messages in enums; dynamic data in factories; invalid data in static. Exactly one right home.
  2. Hard-stop forbidden patterns — the Constitution table triggers refusal, with a concrete alternative each time.
  3. Mandatory exploration — playwright-cli for UI, OpenAPI for API. No guessing selectors or endpoints.
  4. Strict folder discipline — every artifact has one home, so keyword→skill routing works.
  5. Phased skills — you follow the skill's phases instead of inventing a workflow per task.

How a skill loads itself

Describe the work; the routing table (CLAUDE.md §4) picks the first skill, which chains to others. You rarely name a skill — naming one is the fallback when the wrong one loaded.

Day-one

  1. If PROJECT.md is missing, run bootstrap. 2. Read PROJECT.md. 3. Skim the skill for whatever you'll touch first (usually page-objects or api-testing). 4. State your goal; let the skill load and propose; approve / redirect. That loop is the whole job.

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.