agents-united 0.5.0

package/registry/rules/AGENTS.md

@@ -0,0 +1,41 @@
+# Master Operational Rules for Agents United
+
+## 1. Test-Driven Development (TDD)
+- **Mandatory Red-Green-Refactor**: Write failing unit or E2E tests before implementing functional logic or bug fixes.
+- **Deterministic Testing**: Never use arbitrary sleeps or timeouts (`setTimeout`); use auto-waiting assertions and deterministic mock factories.
+- **4-Tier Test Coverage**:
+  - *Tier 1 (Feature Coverage)*: Core function execution and expected outputs.
+  - *Tier 2 (Boundary & Corner Cases)*: Empty inputs, malformed data, network interruptions, and error recovery.
+  - *Tier 3 (Cross-Feature Pairwise)*: Module interoperability between Registry, Installer, Adapters, and CLI.
+  - *Tier 4 (Full Real-World Scenarios)*: Complete inventory audits across all agents, skills, and workflows.
+
+## 2. Code Quality, Modularity & Type Safety
+- **Strict Typing**: Never use implicit or explicit `any` where structured TypeScript types, enums, or interfaces can be declared.
+- **Single Source of Truth**: Centralize shared interfaces in `src/core/types.ts` and maintain schema integrity.
+- **Seams & Decoupling**: Separate business logic, file system operations, and terminal presentation into modular layers.
+
+## 3. Safety, Security & Git Guardrails
+- **Zero Secret Exposure**: Never log, output, or commit API keys, authentication tokens, or `.env` files.
+- **Protected Branch Guard**: Never commit directly to `main`, `master`, or release branches. Always work on feature/fix branches.
+- **Zero Force-Pushes**: Disallow `git push --force` or `git push -f` under all circumstances.
+- **Pre-Staging Diff Inspection**: Always audit staged diffs (`git diff --cached`) before committing.
+
+## 4. Skill Attribution & Provenance Standards
+- **Frontmatter Metadata**: All adopted or adapted skills must declare author metadata (`author`, `version`, `source`, `license`) in YAML frontmatter.
+- **README Credits**: Credit open-source creators and original repositories under `## Credits & Acknowledgments` in `README.md`.
+- **Adaptation Standard**: Adapt external skills to conform to Agents United schema and execution guidelines rather than importing raw unverified files.
+
+## 5. UI/UX Quality, Aesthetics & Accessibility
+- **Terminal Presentation**: Format listings and menus with structured folder tree branches (`├──`, `└──`, `│`), clear badges, and progressive two-stage drill-downs.
+- **Web & Mobile Accessibility**: Enforce WCAG 2.1 Level AA compliance, 4.5:1 text contrast ratios, semantic HTML5, and touch target minimums (44pt iOS, 48dp Android).
+- **Anti-Cliché Design**: Avoid unpadded layouts, purple-on-dark tropes, arbitrary animations, or icon-stuffed bento boxes without functional hierarchy.
+
+## 6. Multi-Agent Delegation & Lifecycle Hooks
+- **Orchestrator Mandate**: Orchestrators plan and coordinate; specialized sub-agents execute domain-specific tasks.
+- **Lifecycle Hooks**: Enforce `PreInvocation`, `PostInvocation`, `PreToolUse`, and `PostToolUse` safety hooks to validate workspace state and audit outputs.
+- **Structured Deliverables**: Provide complete, actionable code and structured markdown summaries with deterministic verification criteria.
+
+## 7. Domain Modeling & Architecture Decision Records (ADRs)
+- **Ubiquitous Language**: Maintain domain terminology consistency across agents, prompts, and documentation.
+- **ADR Governance**: Document non-trivial architectural changes and tradeoffs in `docs/adr/` with Context, Decision, and Consequences.
+- **Socratic Clarification**: Clarify ambiguous requirements via interactive questioning before committing to major architectural shifts.

package/registry/rules/CLAUDE.md

@@ -0,0 +1,12 @@
+# Operational Directives for Anthropic Claude Code
+
+This rule file is automatically provisioned by Agents United to guide Anthropic Claude Code runtime execution.
+
+## Core Behavioral Directives
+
+1. **Test-Driven Development**: Always author failing tests before implementing or refactoring source code.
+2. **Deterministic Verification**: Verify code changes by running test suites (`npm test`) or build steps (`npm run build`) before declaring completion.
+3. **Safety & Zero-Trust Secret Policy**: Never output secret keys or stage `.env` files.
+4. **Git Protection**: Never commit directly to `main` or `master`; do not execute force-pushes (`git push -f`).
+5. **Attribution Standard**: Ensure all adopted skills record author metadata in frontmatter and are acknowledged in `README.md`.
+6. **Clean TypeScript**: Enforce explicit typing and avoid `any`.

package/registry/rules/CURSOR.md

@@ -0,0 +1,12 @@
+# Operational Directives for Cursor & Codex Runtimes
+
+This rule file is automatically provisioned by Agents United to guide Cursor and Codex runtime execution.
+
+## Core Behavioral Directives
+
+1. **Test-Driven First**: Write tests that fail before writing implementation code.
+2. **Deterministic Verification**: Verify code changes with automated test execution before finalizing edits.
+3. **Safety & Zero-Trust Secret Policy**: Never log or commit secret tokens or credentials.
+4. **Git Protection**: Never commit directly to `main` or `master`; zero force-pushes allowed.
+5. **Attribution Standard**: Declare author metadata in skill frontmatter and credit authors in `README.md`.
+6. **Clean Architecture**: Maintain modular boundaries, explicit TypeScript interfaces, and single responsibility.

package/registry/rules/GEMINI.md

@@ -0,0 +1,41 @@
+# Master Operational Rules for Agents United (Gemini / Antigravity Edition)
+
+## 1. Test-Driven Development (TDD)
+- **Mandatory Red-Green-Refactor**: Write failing unit or E2E tests before implementing functional logic or bug fixes.
+- **Deterministic Testing**: Never use arbitrary sleeps or timeouts (`setTimeout`); use auto-waiting assertions and deterministic mock factories.
+- **4-Tier Test Coverage**:
+  - *Tier 1 (Feature Coverage)*: Core function execution and expected outputs.
+  - *Tier 2 (Boundary & Corner Cases)*: Empty inputs, malformed data, network interruptions, and error recovery.
+  - *Tier 3 (Cross-Feature Pairwise)*: Module interoperability between Registry, Installer, Adapters, and CLI.
+  - *Tier 4 (Full Real-World Scenarios)*: Complete inventory audits across all agents, skills, and workflows.
+
+## 2. Code Quality, Modularity & Type Safety
+- **Strict Typing**: Never use implicit or explicit `any` where structured TypeScript types, enums, or interfaces can be declared.
+- **Single Source of Truth**: Centralize shared interfaces in `src/core/types.ts` and maintain schema integrity.
+- **Seams & Decoupling**: Separate business logic, file system operations, and terminal presentation into modular layers.
+
+## 3. Safety, Security & Git Guardrails
+- **Zero Secret Exposure**: Never log, output, or commit API keys, authentication tokens, or `.env` files.
+- **Protected Branch Guard**: Never commit directly to `main`, `master`, or release branches. Always work on feature/fix branches.
+- **Zero Force-Pushes**: Disallow `git push --force` or `git push -f` under all circumstances.
+- **Pre-Staging Diff Inspection**: Always audit staged diffs (`git diff --cached`) before committing.
+
+## 4. Skill Attribution & Provenance Standards
+- **Frontmatter Metadata**: All adopted or adapted skills must declare author metadata (`author`, `version`, `source`, `license`) in YAML frontmatter.
+- **README Credits**: Credit open-source creators and original repositories under `## Credits & Acknowledgments` in `README.md`.
+- **Adaptation Standard**: Adapt external skills to conform to Agents United schema and execution guidelines rather than importing raw unverified files.
+
+## 5. UI/UX Quality, Aesthetics & Accessibility
+- **Terminal Presentation**: Format listings and menus with structured folder tree branches (`├──`, `└──`, `│`), clear badges, and progressive two-stage drill-downs.
+- **Web & Mobile Accessibility**: Enforce WCAG 2.1 Level AA compliance, 4.5:1 text contrast ratios, semantic HTML5, and touch target minimums (44pt iOS, 48dp Android).
+- **Anti-Cliché Design**: Avoid unpadded layouts, purple-on-dark tropes, arbitrary animations, or icon-stuffed bento boxes without functional hierarchy.
+
+## 6. Multi-Agent Delegation & Lifecycle Hooks
+- **Orchestrator Mandate**: Orchestrators plan and coordinate; specialized sub-agents execute domain-specific tasks.
+- **Lifecycle Hooks**: Enforce `PreInvocation`, `PostInvocation`, `PreToolUse`, and `PostToolUse` safety hooks to validate workspace state and audit outputs.
+- **Structured Deliverables**: Provide complete, actionable code and structured markdown summaries with deterministic verification criteria.
+
+## 7. Domain Modeling & Architecture Decision Records (ADRs)
+- **Ubiquitous Language**: Maintain domain terminology consistency across agents, prompts, and documentation.
+- **ADR Governance**: Document non-trivial architectural changes and tradeoffs in `docs/adr/` with Context, Decision, and Consequences.
+- **Socratic Clarification**: Clarify ambiguous requirements via interactive questioning before committing to major architectural shifts.

package/registry/rules/clean-code-and-architecture.md

@@ -0,0 +1,25 @@
+# Persistent Rule: Clean Code & Modular Architecture
+
+## Purpose & Scope
+This rule establishes rigorous standards for code organization, type safety, dependency boundaries, and maintainability across the Agents United ecosystem.
+
+---
+
+## 1. Strict TypeScript & Type Safety
+- **No Implicit `any`**: All function parameters, return values, and exported interfaces must have explicit type declarations.
+- **Centralized Types**: Define and export all shared data models in `src/core/types.ts`.
+- **Exhaustive Pattern Matching**: Use TypeScript discriminated unions for polymorphic types (e.g. `InstallScope`, `InstallMethod`, `AgentHost`).
+
+---
+
+## 2. Modular Architecture & Clean Seams
+- **Single Responsibility Principle**: Keep classes and modules focused on a single domain (e.g. `RegistryResolver` manages assets, `InstallEngine` coordinates installation, `FileAdapter` handles host-specific file writes).
+- **Decoupled Interfaces**: Business logic must never directly invoke terminal I/O (`console.log`, `@clack/prompts`). All user interaction belongs in `src/cli.ts`.
+- **Recursive Bundle Composition**: Support parent-child bundle inheritance (`parentBundle: "software-engineering"`) to eliminate duplicate asset declarations.
+
+---
+
+## 3. Idempotency & Reversibility
+- **Safe State Tracking**: Every installed asset must be recorded in `agents-united.json` with its content hash and installation mode (`symlink` vs `copy`).
+- **Clean Uninstallations**: Uninstallation routines must remove only tracked, managed assets without damaging user-created workspace files.
+- **Zero Orphaned Files**: Installation failures must roll back cleanly or report exact error diagnostic paths.

package/registry/rules/domain-modeling-and-adr.md

@@ -0,0 +1,27 @@
+# Persistent Rule: Domain Modeling & Architecture Decision Records (ADRs)
+
+## Purpose & Scope
+This rule establishes rigorous standards for maintaining a shared ubiquitous language, documenting architectural decisions, and conducting Socratic requirement clarification across all projects.
+
+---
+
+## 1. Ubiquitous Language & Domain Dictionary (`CONTEXT.md`)
+- **Shared Terminology**: Maintain a persistent `CONTEXT.md` defining core domain concepts, entities, and bounded contexts.
+- **Terminology Alignment**: All agents, prompts, APIs, and documentation must use identical terms for domain concepts (e.g. `Bundle`, `Skill`, `Workflow`, `Host`, `Scope`).
+
+---
+
+## 2. Architecture Decision Records (`docs/adr/`)
+Whenever a significant architectural, structural, or strategic decision is made, author an ADR in `docs/adr/`:
+- **File Naming**: `docs/adr/YYYY-MM-DD-<slug>.md` (e.g. `2026-08-14-devops-vs-sysops-separation.md`).
+- **Standard Structure**:
+  - **Status**: Proposed / Accepted / Superseded
+  - **Context**: Problem statement, constraints, and business context.
+  - **Decision**: The selected architectural approach and rationale.
+  - **Consequences**: Positive tradeoffs, limitations, and future implications.
+
+---
+
+## 3. Socratic Requirements Clarification (`/grill-with-docs`)
+- **Probe Ambiguity First**: Before commencing large architectural changes or complex implementations, probe underspecified requirements with structured multiple-choice questions.
+- **Continuous Documentation**: Capture clarified decisions directly into ADRs or domain models during the requirements alignment phase.

package/registry/rules/git-guardrails.md

@@ -0,0 +1,19 @@
+# Persistent Rule: Git Guardrails & Safety Policy
+
+This persistent rule applies across all agent ecosystems and orchestrator bundles. Agents United guarantees user codebase safety by enforcing strict version control guardrails.
+
+## Safety Directives
+
+1. **Never Commit Directly to Protected Branches**:
+   - Always verify the active branch with `git branch --show-current`.
+   - Never commit directly to `main`, `master`, `production`, or `release/*`. Always create a feature or fix branch (`git checkout -b feature/<name>`).
+
+2. **Never Execute Force Pushes**:
+   - Disallow `git push --force` or `git push -f` under all circumstances.
+
+3. **Pre-Staging Secret Scanning**:
+   - Inspect staged changes with `git diff --cached` prior to creating commits.
+   - Never stage `.env`, private keys, API credentials, or auth tokens.
+
+4. **Clean Working Tree**:
+   - Never leave uncommitted temporary files or dirty detached HEAD states.

package/registry/rules/multi-agent-coordination.md

@@ -0,0 +1,33 @@
+# Persistent Rule: Multi-Agent Delegation & Coordination Protocol
+
+## Purpose & Scope
+This rule governs interactions, task distribution, lifecycle hooks, and reporting standards between Lead Orchestrators and specialized sub-agents.
+
+---
+
+## 1. Separation of Responsibilities
+- **Lead Orchestrators**:
+  - Analyze user requirements and formulate step-by-step implementation plans.
+  - Route tasks to specialized sub-agents via structured prompts.
+  - Synthesize sub-agent outputs and enforce verification phase gates.
+- **Specialized Sub-Agents**:
+  - Execute focused domain tasks (e.g. `subagent-ios-architect`, `subagent-data-engineer`, `subagent-qa-automation-lead`).
+  - Report findings, code modifications, and test results back to the caller without executing unauthorized out-of-scope work.
+
+---
+
+## 2. Lifecycle Hook Governance
+All agent markdown manifests must support declarative lifecycle hooks:
+- **`PreInvocation`**: Executed before task execution begins to verify preconditions, workspace paths, and active branch safety.
+- **`PostInvocation`**: Executed upon task completion to verify test pass rates, lint status, and artifact generation.
+- **`PreToolUse`**: Guards sensitive commands against destructive or unconfirmed executions.
+- **`PostToolUse`**: Validates tool output integrity and logs audit trails.
+
+---
+
+## 3. Structured Deliverable Standard
+Sub-agents must provide structured outputs containing:
+1. **Executive Summary**: 1-2 sentence overview of actions taken.
+2. **File Modifications**: Clickable file links and concise change summaries.
+3. **Verification Results**: Command output and test pass indicators.
+4. **Next Actions / Risks**: Open decisions requiring human or orchestrator review.

package/registry/rules/quality-aesthetics-accessibility.md

@@ -0,0 +1,29 @@
+# Persistent Rule: Quality, Aesthetics & Accessibility Standards
+
+## Purpose & Scope
+This rule mandates premium visual design, accessible human-computer interaction, and intuitive terminal user interfaces across all tools and web components produced by Agents United.
+
+---
+
+## 1. Terminal User Interface (TUI) Standards
+- **Folder Tree Visualizations**: Present multi-item catalogs using standardized Unicode box-drawing characters (`├──`, `└──`, `│`, `│   `).
+- **Two-Stage Progressive Drill-Downs**: Avoid massive unorganized select lists. Group options into domain categories before opening specialized sub-teams.
+- **Consistent Visual Badges**: Use color-coded badges to indicate item types (`[Bundle]`, `[Agent]`, `[Skill]`, `[Workflow]`).
+- **Explicit Confirmation Gates**: Require user confirmation before destructive actions or large-scale batch installations (e.g. installing an entire department).
+
+---
+
+## 2. Web & Mobile Accessibility Standards (WCAG 2.1 Level AA)
+- **Semantic HTML5**: Always use appropriate semantic elements (`<main>`, `<nav>`, `<article>`, `<header>`, `<button>`) instead of nested `<div>` wrappers.
+- **Contrast Ratios**: Maintain a minimum contrast ratio of 4.5:1 for normal text and 3:1 for large text and interactive components.
+- **Keyboard Navigation**: Guarantee visible focus rings and logical tab ordering (`tabindex="0"` for interactive custom widgets).
+- **Touch Target Ergonomics**: Ensure minimum interactive target sizes of 44x44 points (iOS) and 48x48 dp (Android).
+
+---
+
+## 3. Anti-Cliché Design Guidelines
+Unless explicitly requested by the user, DO NOT use generic tropes:
+- No purple fonts or violet accents on dark backgrounds.
+- No unpadded containers or textureless surfaces.
+- No floating badge pills with pulsing dots above main headlines.
+- No icon-stuffed bento boxes without clear functional hierarchy.

package/registry/rules/skill-attribution.md

@@ -0,0 +1,27 @@
+# Persistent Rule: Skill & Workflows Author Attribution Standard
+
+This rule governs the adoption, adaptation, and integration of external skills, agents, workflows, and prompts into the Agents United ecosystem.
+
+## Mandatory Attribution Policy
+
+Whenever skills or workflows are adopted or adapted from external creators or repositories, you MUST:
+
+1. **Skill Frontmatter Metadata**:
+   - In `SKILL.md`, include the author's name, repository URL, and license inside the YAML frontmatter `metadata` block:
+   ```yaml
+   ---
+   name: <skill-name>
+   description: <skill-description>
+   metadata:
+     author: "<Author Name> (<repo-identifier>)"
+     version: "1.0.0"
+     source: "https://github.com/<owner>/<repo>"
+     license: "MIT"
+   ---
+   ```
+
+2. **README.md Credits & Acknowledgments Section**:
+   - Add or maintain an entry in `README.md` under `## Credits & Acknowledgments` with the author's handle, repository link, and list of adapted capabilities.
+
+3. **Integrity & Adaptation**:
+   - Adapt external skills to conform to Agents United schema and execution standards rather than importing them as unformatted raw files.

package/registry/rules/test-driven-development.md

@@ -0,0 +1,41 @@
+# Persistent Rule: Test-Driven Development (TDD) Protocol
+
+## Purpose & Scope
+This rule enforces a strict, verifiable Test-Driven Development protocol across all software engineering, architecture, and tool development tasks within Agents United.
+
+---
+
+## 1. The Red-Green-Refactor Lifecycle
+Autonomous agents must follow the three-phase cycle for all new features, refactors, and bug fixes:
+1. **Red Phase (Write Failing Tests)**: Create or update test files specifying the expected behavior, types, or assertions before modifying implementation source code. Execute `npm test` and verify that the tests fail for the intended reason.
+2. **Green Phase (Minimal Implementation)**: Write only the minimal necessary implementation code to satisfy the test assertions and achieve a green test pass.
+3. **Refactor Phase (Clean Up & Optimize)**: Refactor code for clarity, performance, and modularity while ensuring all tests continue to pass.
+
+---
+
+## 2. The 4-Tier Testing Methodology
+All test suites in the codebase should adhere to our 4-tier architectural testing model:
+
+### Tier 1: Feature Coverage (Happy Path)
+- Verify that standard valid inputs produce expected outputs.
+- Test all exposed functions, classes, and exported interface methods.
+
+### Tier 2: Boundary & Corner Cases (Negative Testing)
+- Test empty strings, null/undefined inputs, and invalid type parameters.
+- Test missing files, malformed YAML/JSON, and invalid enum values.
+- Verify graceful error messages and structured exception handling.
+
+### Tier 3: Cross-Feature Integration (Pairwise Consistency)
+- Verify data flow across interconnected modules (e.g. `RegistryResolver` $\leftrightarrow$ `InstallEngine` $\leftrightarrow$ `FileAdapter`).
+- Ensure lockfile updates match physical file system states.
+
+### Tier 4: Real-World Scenarios & Full Inventory Audits
+- Execute end-to-end tests against the complete registry catalog (all agents, skills, workflows).
+- Verify zero schema drift, zero duplicate naming, and valid frontmatter parsing.
+
+---
+
+## 3. Flakiness Elimination Rules
+- **No Hardcoded Sleeps**: Never use `setTimeout` or arbitrary delays. Always use auto-waiting assertions (`expect(locator).toBeVisible()`) or deterministic callbacks.
+- **Isolated Workspaces**: Run test file mutations inside temporary directories (e.g. `scratch/test-workspace/`) and clean up in `afterEach()`.
+- **Deterministic Factories**: Use predictable test fixtures rather than non-deterministic random data.

package/registry/skills/ab-test-setup/SKILL.md

@@ -0,0 +1,150 @@
+---
+name: ab-test-setup
+description: Production-grade Ab Test Setup playbook for design operations, UX
+  systems, and growth strategy.
+metadata:
+  author: agents-united
+  version: 2.0.0
+  icon: 📊
+disable-slash-command: true
+---
+
+# Ab Test Setup
+
+## Overview & Purpose
+The Ab Test Setup skill provides a deterministic framework for executing ab test setup processes in modern software products.
+
+Following this skill ensures high usability, visual consistency, rapid iteration, and complete cross-functional team alignment.
+
+## Execution Triggers & Prerequisites
+### Execution Triggers
+- Direct request to execute Ab Test Setup tasks.
+- Auditing existing product assets or workflows.
+- Standardizing ab-test-setup procedures across team projects.
+- Preparing design handoffs or growth campaign launches.
+
+### Prerequisites
+- Project workspace configured with design system tokens or component libraries.
+- Target UI design specification or growth experiment hypothesis.
+- Testing and linting tools operational.
+- Clean git working directory.
+
+## Input & Output Requirements
+### Inputs
+| Parameter | Type | Required | Description |
+|---|---|---|---|
+| `target_scope` | String | Yes | Target UI component, page, or campaign scope |
+| `config` | Object | Optional | Specific parameters and threshold configurations |
+| `output_dir` | Directory Path | Optional | Destination directory for generated artifacts |
+| `strict_mode` | Boolean | Optional | Enforce strict zero-warning validation |
+
+### Outputs
+| Artifact | Path / Format | Description |
+|---|---|---|
+| Specification Document | `docs/ab-test-setup/spec.md` | Full specification and guidelines document |
+| Component / Asset Files | `src/ab-test-setup/*` | Implemented design tokens, components, or campaign assets |
+| Audit Report | `reports/ab-test-setup/summary.json` | Health check and audit metric results |
+
+## Step-by-Step Execution Runbook
+
+### Phase 1: Pre-Execution Discovery & Workspace Analysis
+1. Inspect workspace repository to locate relevant UI components, tokens, or campaign assets.
+   ```bash
+   find src/ docs/ -maxdepth 3 -type f
+   ```
+2. Analyze domain requirements and classify core UI elements, interaction flows, or growth metrics.
+3. Establish baseline quality metrics and target benchmarks.
+4. Verify working tree status to ensure clean git workspace.
+   ```bash
+   git status --short
+   ```
+5. Formulate initial execution plan.
+
+### Phase 2: Input Contract Validation & Strategy Selection
+1. Validate input parameters against technical feasibility and design system guidelines.
+2. Select implementation pattern matching component or campaign architecture.
+3. Establish verification rules and accessibility / conversion thresholds.
+4. Formulate atomic step-by-step execution sequence.
+5. Create temporary working directory if needed.
+
+### Phase 3: Core Step-by-Step Implementation Execution
+1. Author primary specification document at `docs/ab-test-setup/spec.md`.
+2. Generate code, token, or layout implementation files.
+   ```bash
+   npm run typecheck
+   ```
+3. Apply automated formatting and linting tools.
+4. Execute unit or visual regression tests.
+   ```bash
+   npm test
+   ```
+5. Refactor asset structure for optimal performance and maintainability.
+
+### Phase 4: Verification, Testing & Quality Gate Checking
+1. Run full project verification suite.
+   ```bash
+   npm run typecheck && npm test && npm run build
+   ```
+2. Verify zero lint errors, type warnings, or broken references.
+3. Execute CLI health doctor check.
+   ```bash
+   node dist/cli.js doctor
+   ```
+4. Assert all acceptance criteria are satisfied.
+
+### Phase 5: Post-Execution Cleanup & Artifact Generation
+1. Generate execution summary report at `reports/ab-test-setup/summary.md`.
+2. Clean up temporary build artifacts and scratch files.
+3. Commit generated files to git repository.
+   ```bash
+   git add docs/ab-test-setup/ reports/ab-test-setup/
+   git commit -m "feat(ab-test-setup): implement Ab Test Setup playbook artifacts"
+   ```
+4. Publish documentation for team review.
+
+## Code & Configuration Exemplars
+
+### Exemplar 1: Ab Test Setup Configuration Specification
+```yaml
+version: "2.0.0"
+metadata:
+  skill: "ab-test-setup"
+  author: "agents-united"
+rules:
+  strictValidation: true
+  reporting:
+    format: "json"
+    output: "reports/ab-test-setup/summary.json"
+```
+
+### Exemplar 2: Ab Test Setup Helper Module
+```typescript
+export function runAbTestSetup(scope: string): boolean {
+  console.log('Running Ab Test Setup on:', scope);
+  return true;
+}
+```
+
+## Edge Cases & Error Recovery Procedures
+
+### Scenario A: Validation Failure in Ab Test Setup
+1. **Diagnosis**: Specification or code asset fails validation rules in ab-test-setup.
+2. **Recovery Protocol**:
+   - Step 1: Inspect error log at reports directory.
+   - Step 2: Correct non-compliant syntax or structure.
+   - Step 3: Re-run verification pipeline.
+
+### Scenario B: Missing Resource for Ab Test Setup
+1. **Diagnosis**: Target design token or configuration asset missing from workspace.
+2. **Recovery Protocol**:
+   - Step 1: Generate baseline resource file from standard template.
+   - Step 2: Update configuration references.
+   - Step 3: Resume runbook execution.
+
+## Verification & Validation Checklist
+- [ ] Frontmatter conforms strictly to `author: "agents-united"` and `version: "2.0.0"`.
+- [ ] All 7 mandatory sections present with explicit headers.
+- [ ] Step-by-Step Execution Runbook body contains >= 50 lines.
+- [ ] Code exemplars provided with valid syntax fencing.
+- [ ] Zero dummy placeholder strings or unpopulated template markers present.
+- [ ] Project build, test suite, and doctor check pass 100% cleanly.

package/registry/skills/accessibility-audit/SKILL.md

@@ -0,0 +1,150 @@
+---
+name: accessibility-audit
+description: Production-grade Accessibility Audit playbook for design
+  operations, UX systems, and growth strategy.
+metadata:
+  author: agents-united
+  version: 2.0.0
+  icon: ♿
+disable-slash-command: true
+---
+
+# Accessibility Audit
+
+## Overview & Purpose
+The Accessibility Audit skill provides a deterministic framework for executing accessibility audit processes in modern software products.
+
+Following this skill ensures high usability, visual consistency, rapid iteration, and complete cross-functional team alignment.
+
+## Execution Triggers & Prerequisites
+### Execution Triggers
+- Direct request to execute Accessibility Audit tasks.
+- Auditing existing product assets or workflows.
+- Standardizing accessibility-audit procedures across team projects.
+- Preparing design handoffs or growth campaign launches.
+
+### Prerequisites
+- Project workspace configured with design system tokens or component libraries.
+- Target UI design specification or growth experiment hypothesis.
+- Testing and linting tools operational.
+- Clean git working directory.
+
+## Input & Output Requirements
+### Inputs
+| Parameter | Type | Required | Description |
+|---|---|---|---|
+| `target_scope` | String | Yes | Target UI component, page, or campaign scope |
+| `config` | Object | Optional | Specific parameters and threshold configurations |
+| `output_dir` | Directory Path | Optional | Destination directory for generated artifacts |
+| `strict_mode` | Boolean | Optional | Enforce strict zero-warning validation |
+
+### Outputs
+| Artifact | Path / Format | Description |
+|---|---|---|
+| Specification Document | `docs/accessibility-audit/spec.md` | Full specification and guidelines document |
+| Component / Asset Files | `src/accessibility-audit/*` | Implemented design tokens, components, or campaign assets |
+| Audit Report | `reports/accessibility-audit/summary.json` | Health check and audit metric results |
+
+## Step-by-Step Execution Runbook
+
+### Phase 1: Pre-Execution Discovery & Workspace Analysis
+1. Inspect workspace repository to locate relevant UI components, tokens, or campaign assets.
+   ```bash
+   find src/ docs/ -maxdepth 3 -type f
+   ```
+2. Analyze domain requirements and classify core UI elements, interaction flows, or growth metrics.
+3. Establish baseline quality metrics and target benchmarks.
+4. Verify working tree status to ensure clean git workspace.
+   ```bash
+   git status --short
+   ```
+5. Formulate initial execution plan.
+
+### Phase 2: Input Contract Validation & Strategy Selection
+1. Validate input parameters against technical feasibility and design system guidelines.
+2. Select implementation pattern matching component or campaign architecture.
+3. Establish verification rules and accessibility / conversion thresholds.
+4. Formulate atomic step-by-step execution sequence.
+5. Create temporary working directory if needed.
+
+### Phase 3: Core Step-by-Step Implementation Execution
+1. Author primary specification document at `docs/accessibility-audit/spec.md`.
+2. Generate code, token, or layout implementation files.
+   ```bash
+   npm run typecheck
+   ```
+3. Apply automated formatting and linting tools.
+4. Execute unit or visual regression tests.
+   ```bash
+   npm test
+   ```
+5. Refactor asset structure for optimal performance and maintainability.
+
+### Phase 4: Verification, Testing & Quality Gate Checking
+1. Run full project verification suite.
+   ```bash
+   npm run typecheck && npm test && npm run build
+   ```
+2. Verify zero lint errors, type warnings, or broken references.
+3. Execute CLI health doctor check.
+   ```bash
+   node dist/cli.js doctor
+   ```
+4. Assert all acceptance criteria are satisfied.
+
+### Phase 5: Post-Execution Cleanup & Artifact Generation
+1. Generate execution summary report at `reports/accessibility-audit/summary.md`.
+2. Clean up temporary build artifacts and scratch files.
+3. Commit generated files to git repository.
+   ```bash
+   git add docs/accessibility-audit/ reports/accessibility-audit/
+   git commit -m "feat(accessibility-audit): implement Accessibility Audit playbook artifacts"
+   ```
+4. Publish documentation for team review.
+
+## Code & Configuration Exemplars
+
+### Exemplar 1: Accessibility Audit Configuration Specification
+```yaml
+version: "2.0.0"
+metadata:
+  skill: "accessibility-audit"
+  author: "agents-united"
+rules:
+  strictValidation: true
+  reporting:
+    format: "json"
+    output: "reports/accessibility-audit/summary.json"
+```
+
+### Exemplar 2: Accessibility Audit Helper Module
+```typescript
+export function runAccessibilityAudit(scope: string): boolean {
+  console.log('Running Accessibility Audit on:', scope);
+  return true;
+}
+```
+
+## Edge Cases & Error Recovery Procedures
+
+### Scenario A: Validation Failure in Accessibility Audit
+1. **Diagnosis**: Specification or code asset fails validation rules in accessibility-audit.
+2. **Recovery Protocol**:
+   - Step 1: Inspect error log at reports directory.
+   - Step 2: Correct non-compliant syntax or structure.
+   - Step 3: Re-run verification pipeline.
+
+### Scenario B: Missing Resource for Accessibility Audit
+1. **Diagnosis**: Target design token or configuration asset missing from workspace.
+2. **Recovery Protocol**:
+   - Step 1: Generate baseline resource file from standard template.
+   - Step 2: Update configuration references.
+   - Step 3: Resume runbook execution.
+
+## Verification & Validation Checklist
+- [ ] Frontmatter conforms strictly to `author: "agents-united"` and `version: "2.0.0"`.
+- [ ] All 7 mandatory sections present with explicit headers.
+- [ ] Step-by-Step Execution Runbook body contains >= 50 lines.
+- [ ] Code exemplars provided with valid syntax fencing.
+- [ ] Zero dummy placeholder strings or unpopulated template markers present.
+- [ ] Project build, test suite, and doctor check pass 100% cleanly.