agentsclimarketplace

Daa review

Skill gigayaya/DAA-Master/plugins/DAA-Master/skills/daa-review

An agent Skills plugin for the Declarative Action Architecture (DAA). This plugin transforms Claude into an expert in the Declarative Action Architecture (DAA), offering four core capabilities.

Install
npx -y skills add gigayaya/DAA-Master --skill daa-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when reviewing existing test code for DAA compliance, identifying anti-patterns, or suggesting improvements to automation test architecture

SKILL.md

3.6 KB, as published. Nobody here has run it

DAA Code Reviewer

REQUIRED BACKGROUND: You MUST understand daa:daa-core before using this skill.

Overview

Review automation test code against DAA principles. Classify each violation by severity and provide actionable fix suggestions.

Review Process

digraph review {
    rankdir=TB;
    "Read the code" [shape=box];
    "Classify into layers" [shape=box];
    "Check Test Layer rules" [shape=box];
    "Check Action Layer rules" [shape=box];
    "Check Physical Layer rules" [shape=box];
    "Check cross-layer boundaries" [shape=box];
    "Assign severity" [shape=box];
    "Report findings" [shape=doublecircle];

    "Read the code" -> "Classify into layers";
    "Classify into layers" -> "Check Test Layer rules";
    "Check Test Layer rules" -> "Check Action Layer rules";
    "Check Action Layer rules" -> "Check Physical Layer rules";
    "Check Physical Layer rules" -> "Check cross-layer boundaries";
    "Check cross-layer boundaries" -> "Assign severity";
    "Assign severity" -> "Report findings";
}

Quick Review Checklist (Top 10)

Run through these checks in order. Stop and flag immediately on any CRITICAL finding.

Test Layer

  1. [CRITICAL] Does any test method contain if, for, while, or try/catch?
  2. [CRITICAL] Does any test method make direct API/UI/DB calls (bypassing Action Layer)?
  3. [WARNING] Does any test method contain assert statements (should be in Action Layer)?

Action Layer

  1. [CRITICAL] Does any action method lack self-verification (no assertion after operation)?
  2. [CRITICAL] Does any action method call the underlying library directly (bypassing Physical Layer)?
  3. [WARNING] Are there Composite Actions that call Physical Layer directly instead of composing Atomics?
  4. [WARNING] Do action names follow the verb_object_and_verify_outcome pattern?

Physical Layer

  1. [CRITICAL] Does the Physical Layer contain any assertions or business logic?
  2. [WARNING] Does any Physical Layer method perform multiple operations instead of one?

Cross-Layer

  1. [CRITICAL] Does any layer skip the adjacent layer (e.g., Test → Physical directly)?

→ Full detailed checklist: checklist.md

Severity Classification

LevelMeaningAction Required
CRITICALBreaks DAA fundamentals — causes false positives or destroys test trustMust fix before merge
WARNINGHurts maintainability or violates DAA best practicesShould fix; acceptable to defer with justification
SUGGESTIONImprovement opportunity for readability or consistencyNice to have; fix when convenient

→ Full severity guide with examples: severity-guide.md

Report Format

Every report MUST start with a DAA Score (1-10, where 10 = full compliance). Score is calculated by deducting from 10 based on findings severity.

→ Full scoring rubric, rules, and report template: scoring.md

Common Patterns to Watch For

  1. "It works so it's fine": Code that passes tests but violates DAA is technical debt that will cause false positives later
  2. Gradual erosion: One if in a test method → two → tests become procedural scripts
  3. "Just this once": Direct Physical Layer calls in tests "just for this special case" — there are no exceptions
  4. Assertion-free actions: Methods named click_save() without _and_verify_* suffix — likely missing verification

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.