One-click install
npx skills add https://github.com/redhat-community-ai-tools/UnifAI --skill pipeline-design-reviewer
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pipeline-design-reviewer
Source: https://github.com/redhat-community-ai-tools/UnifAI/tree/main/.cursor/skills/pipeline-design-reviewer
Command: npx skills add https://github.com/redhat-community-ai-tools/UnifAI --skill pipeline-design-reviewer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents costly implementation mistakes by forcing early, critical review of technical pipeline designs for architectural correctness, performance risks, integration gaps, and missing requirements.

Core Features & Use Cases

  • Hexagonal architecture compliance checks: aggressively flags dependency-direction violations and leaky layering as CRITICAL.
  • Efficiency & scalability scrutiny: challenges over-engineering, redundant calls, and likely bottlenecks in the proposed flow.
  • Risk and impact assessment: identifies breaking-change hazards, side effects, and what regression testing will be needed.
  • Mandatory layer completeness verification: ensures the design covers UI, inbound adapters, and any data/seed components when applicable, marking gaps as CRITICAL — INCOMPLETE DESIGN.
  • Auth and external protocol realism: verifies OAuth/MCP discovery, issuer alignment, and credential storage paths; otherwise marks as UNVERIFIED.
  • External dependency failure modes: requires explicit handling for 401/503/timeouts plus clear degradation vs error behavior, marking omissions as CRITICAL.
  • Local dev and partial-access deployment strategy: requires mocks/fakes and fallback wiring so the system still starts when a dependency is missing; omissions are CRITICAL.

Quick Start

Ask the pipeline designer to provide the proposed design document, then run pipeline-design-reviewer to produce a critical, actionable set of fixes before implementation begins.

Frequently Asked Questions about pipeline-design-reviewer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review a pipeline design for hexagonal architecture violations before implementation?

Pipeline design review checks for hexagonal architecture compliance by aggressively flagging dependency-direction violations and leaky layering as critical errors. This enforces proper dependency inversion before any implementation work begins.

What happens if my AI pipeline design is missing external dependency failure handling?

Missing failure handling for external dependencies like 401, 503, or timeouts results in a critical design flag. The review requires explicit degradation versus error behavior to ensure pipeline resilience before implementation.

How do I validate OAuth and MCP protocol references in a technical design document?

Auth and external protocol realism validates OAuth/MCP discovery, issuer alignment, and credential storage paths within the design. Unverifiable auth configurations are marked as unverified to prevent integration failures.

Does pipeline design review check for missing local-dev and partial-access deployment strategies?

Pipeline design review verifies local-dev and partial-access deployment strategies by requiring mocks, fakes, and fallback wiring. Omissions are marked critical to ensure the system starts when external dependencies are missing.

What's the best way to assess efficiency and scalability risks in a multi-agent pipeline design?

Efficiency and scalability scrutiny challenges over-engineering, redundant API calls, and likely bottlenecks in the proposed flow. This identifies breaking-change hazards and determines regression testing needs early.

When do I need a mandatory layer completeness verification for my pipeline architecture?

Mandatory layer completeness verification is needed when a design must cover UI, inbound adapters, and data or seed components. Missing coverage is marked as a critical incomplete design to prevent architectural gaps.