From git-recon
Get a second opinion on your current implementation plan by sending it to Codex for independent review. Use after brainstorming, after writing a plan, or when you want to iterate on a plan before execution. Triggers on 'get a second opinion', 'review plan with codex', 'codex review', or 'iterate on plan'.
How this skill is triggered — by the user, by Claude, or both
Slash command
/git-recon:codex-plan-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Send the current working plan to Codex 5.4 for an independent second opinion. This skill is used to iterate on plans before committing to implementation.
Send the current working plan to Codex 5.4 for an independent second opinion. This skill is used to iterate on plans before committing to implementation.
Locate the most recent plan by checking these locations in order:
docs/superpowers/specs/*.md (most recent by date prefix)docs/superpowers/plans/*.md (most recent by date prefix).claude/memory-bank/activeContext.md for references to current workIf no plan file is found, ask the user to point you to the plan or spec they want reviewed.
Read the full plan file. Construct a prompt for Codex that includes:
Run Codex with these exact settings — non-spark model, high effort, read-only:
codex exec \
-m gpt-5.4-codex \
--config model_reasoning_effort="high" \
--sandbox read-only \
--skip-git-repo-check \
--full-auto \
2>/dev/null \
"<prompt>"
The prompt sent to Codex must follow this template:
You are a senior architect reviewing an implementation plan. Your job is to find weaknesses, missing considerations, and improvements. Be direct and specific.
PROJECT CONTEXT:
<brief project description from CLAUDE.md or memory bank>
PLAN TO REVIEW:
<full plan content>
Review this plan and provide:
1. **Structural Issues** — Are there missing steps, wrong ordering, or circular dependencies?
2. **Architectural Concerns** — Does the plan fight existing patterns? Are there simpler approaches?
3. **Risk Assessment** — What could go wrong? What assumptions are fragile?
4. **Missing Considerations** — What did the plan author overlook? (edge cases, error handling, testing, rollback)
5. **Specific Improvements** — Concrete suggestions, not vague advice. Reference specific plan sections.
Be critical. The goal is to strengthen the plan, not validate it.
After Codex returns its review:
If the user wants to update the plan:
gpt-5.4-codex (not spark, not 5.3)high reasoning effortread-only sandbox — this is a review, not an editcodex resume for follow-up questionsnpx claudepluginhub smirnov-labs/claude-skills --plugin presentation-toolsProvides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.