From jarvis-plan-review
Reviews design plans and RFCs across 11 dimensions including security, architecture, and observability. Pushes plans toward best-in-class with structured pass/risk/block output.
How this skill is triggered — by the user, by Claude, or both
Slash command
/jarvis-plan-review:plan-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Reviews a feature plan, RFC, or architecture doc across 11 independent dimensions. Goal: push the plan to "best-in-class" — not to rubber-stamp it.
Reviews a feature plan, RFC, or architecture doc across 11 independent dimensions. Goal: push the plan to "best-in-class" — not to rubber-stamp it.
For each section, the review returns:
Pass — meets barRisk — works but has identifiable risk → suggested mitigationBlock — critical gap → must address before implementationFor deep analysis, run with --effort xhigh (Opus 4.7 only) to maximize reasoning depth:
claude --effort xhigh
> /plan-review <plan-doc-path>
Cost approximately 2x baseline, but materially improves gap detection — especially in security and reliability sections.
Block, do I have a concrete falsification — what evidence shows this would actually fail?Adapted from gstack (paretofilm/superpowers-gstack) /plan-ceo-review. Section list refined for sole-developer + AI-pair-programming workflows where the "CEO review" framing isn't directly applicable.
/brainstorm — earlier ideation stage/deep-interview — convergent requirement gating (predecessor)/verify — implementation verification (successor)/security-review — focused security audit (deeper than section 4 alone)npx claudepluginhub ramsbaby/jarvis --plugin jarvis-plan-reviewGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.