skills.rest
daudaudinang
daudaudinangCOMMUNITY
Explore all skills by this author →
GitHub

persona-reviewer

Coordinate multi-perspective plan and code reviews with four canonical agents.

Updated Apr 14, 2026
Search Tags:#orchestration#quality-assurance#code-review#architecture-review#agent-based#planning-review#multipersona
One-click install
npx skills add https://github.com/daudaudinang/vibecode --skill persona-reviewer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: persona-reviewer
Source: https://github.com/daudaudinang/vibecode/tree/main/.agents/skills/persona-reviewer
Command: npx skills add https://github.com/daudaudinang/vibecode --skill persona-reviewer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Teams need thorough, unbiased reviews of plans and code from multiple perspectives to improve quality and reduce blind spots.

Core Features & Use Cases

  • Canonical personas available: Senior PM, Senior UI/UX Designer, Senior Developer, System Architecture
  • Modes of orchestration: Standard mode with independent agents, Fast mode for rapid delta review
  • Result consolidation: Orchestrator merges evidence and yields final verdict suitable for decision-making
  • Use case: When evaluating a new feature spec and its implementation, run reviews from four angles, merge findings, and update the plan accordingly.

Quick Start

Invoke the reviewer with a target artifact path to obtain a multi-perspective assessment.

Frequently Asked Questions about persona-reviewer

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

FAQPage Schema
How do I run multi-perspective code and architecture reviews to reduce blind spots?▼

Multi-perspective code and architecture reviews use a four-agent roster—senior PM, UI/UX designer, developer, and system architect—to independently assess artifacts, collect evidence, reconcile conflicts, and output a consolidated verdict.

What is the best way to automate plan review from different team perspectives?▼

Automated plan review uses orchestrator logic to spawn independent agent personas that evaluate feature specs from four distinct angles, merging findings into a final decision-ready verdict.

How do I perform a rapid re-evaluation of updated code without a full review?▼

Rapid re-evaluation uses a fast mode to perform quick delta reviews across the four canonical perspectives, checking only the changed sections rather than the entire codebase.

Can I review both a feature spec and its implementation at the same time?▼

Yes, you can review a feature spec and its implementation simultaneously by invoking the reviewer with the target artifact path to obtain a multi-perspective assessment covering PM, UI/UX, developer, and architecture viewpoints.

Does the multi-persona review process require external dependencies or components?▼

No, the multi-persona review process requires no external dependencies or components, running entirely through its internal orchestrator logic to spawn and manage the independent agents.

How does the orchestrator handle conflicting feedback from different review personas?▼

The orchestrator handles conflicting feedback by collecting evidence from all independent agents, reconciling the conflicts internally, and yielding a single consolidated verdict suitable for decision-making.

Related Skills

Explore More Software Engineering Skills→
axel-kaliffaxel-kaliff

grumpy-review

Review code for bugs in error paths, resource lifecycles, and concurrency.

Community
Intermediate
prestariusprestarius
1

skill-stocktake

Audit Claude Code skills for overlap, freshness, and trigger integrity.

Community
Advanced
samjhechtsamjhecht
4

requesting-code-review

Dispatch a code-review subagent to verify implementation against plan or requirements.

Community
Basic
ZaxbyHubZaxbyHub
437

council

Coordinates three AI agents to jointly evaluate complex questions and produce cited advisory answers.

Official
Advanced