review

You can conduct a multi-persona project assessment by running the review Skill on your SPADES Scope, Plan, or both. It uses four specialized subagents to independently verify design, security, and strategic considerations, producing a merged and prioritized findings report.

Multi-persona quality assurance review uses specialized subagents like scope-guardian, architecture-strategist, security-lens, and adversarial-reviewer to examine a project from diverse perspectives. This mechanism ensures thorough verification of requirements, risks, and strategic alignment by highlighting convergence among the reviewers' opinions.

Yes, you need the SPADES framework and its associated persona agents installed to run this project assessment. The review Skill explicitly depends on the spades-framework to gather context from your architecture documentation and execute the four specialized subagents.

To start a scope review, use the command `/spades:review scope <scope-title>`. To review a plan, use `/spades:review plan <plan-title>`. You can choose between Scope Review, Plan Review, or Full Review modes based on your current project stage and requirements.

The best way to get structured findings is using a multi-persona review that merges and prioritizes subagent opinions into a single report. This approach highlights opinion convergence and provides both a CLI digest and a persisted full report in Markdown or HTML format.

You should not use a multi-persona review if your project lacks established SPADES Scope or Plan documentation. The review process requires context gathering from these specific architecture documents to function correctly and provide meaningful independent verification.