This skill is for Claude only. Codex and Gemini CLI should not invoke it.

Second Opinions

Delegate tasks to OpenAI Codex (MCP) and/or Google Gemini (CLI) for independent perspectives from different model families. Works for code review, plan evaluation, doc editing, codebase analysis, or any arbitrary task.

Modes: both (default), codex, gemini

When To Use

• Code review before committing, opening a PR, or merging
• Plan or architecture review from multiple perspectives
• Documentation review or content editing
• Codebase analysis (patterns, race conditions, dead code)
• Any task benefiting from parallel exploration by external models

Scope

• Skip for trivial tasks you can verify directly
• Skip when the user asked for your own judgment only
• Exclude secrets, credentials, and tokens from prompts

User Instructions

Follow any user instructions below. They override the standard workflow when conflicts arise.

<user_instructions> $ARGUMENTS </user_instructions>

Workflow

1. Gather Context

Use AskUserQuestion to collect parameters. Skip questions whose answers are obvious from user arguments. Combine into one call (max 4 questions).

Tool (always ask unless specified):

header: "Tool"
question: "Which model should run this?"
options:
  - "Both Codex and Gemini (Recommended)" -> parallel dispatch
  - "Codex only"
  - "Gemini only"

Task type (always ask unless clear from arguments):

header: "Task type"
question: "What kind of task?"
options:
  - "Code review (diff-based)"
  - "Plan or architecture review"
  - "Documentation or content review"
  - "Custom task"

Focus (skip for custom tasks where the user already specified intent):

header: "Focus"
question: "Any specific focus?"
options:
  - "General review"
  - "Security & auth"
  - "Performance"
  - "Correctness & edge cases"

Diff scope (only for code review):

header: "Diff scope"
question: "What should be reviewed?"
options:
  - "Uncommitted changes"  -> git diff HEAD + untracked
  - "Branch diff vs main"  -> git diff <branch>...HEAD
  - "Specific commit"      -> follow up for SHA

2. Build Material

Gather content based on task type:

Auto-detect default branch for code reviews:

git symbolic-ref refs/remotes/origin/HEAD 2>/dev/null | sed 's@^refs/remotes/origin/@@' || echo main

3. Construct & Dispatch

Prompt structure — use XML tags for unambiguous parsing. Put long content first, instructions after (per Anthropic long-context best practices):

<material>
[The diff, plan, document, codebase excerpt, or task input — long content goes first]
</material>

<context>
[Project conventions from CLAUDE.md, if it exists]
</context>

<role>
[Role appropriate to the task — e.g., independent code reviewer, architecture evaluator, technical editor]
</role>

<task>
[What to do — review, analyze, evaluate, rewrite, etc.]
</task>

<focus>
[Focus area if specified]
</focus>

<output_format>
Structure your response as:
1. Critical issues (blocking)
2. Important concerns (should address)
3. Minor suggestions (nice to have)
4. What's done well

End with a clear verdict and confidence score (0-1).
</output_format>

Adapt <role>, <task>, and <output_format> to the task type. For code reviews, use the OpenAI cookbook prompt from codex.md. For custom tasks, translate the user's intent directly.

Project context: If CLAUDE.md exists in the repo root, always include it so reviewers check against project conventions.

See references/codex.md for Codex MCP prompt patterns and invocation. See references/gemini.md for Gemini CLI invocation patterns.

Dispatch rules:

• When running both tools, issue mcp__codex__codex and Bash(gemini ...) calls in a single message for parallel execution.
• Dispatch directly without pre-checking tool availability. If a tool fails, run only the available one.

Iterative mode (for plans/architecture, or when explicitly requested):

1. Dispatch initial review
2. If reviewer returns significant concerns, revise the material based on feedback
3. Re-submit using mcp__codex__codex-reply (with threadId) or a fresh gemini call including revision context
4. Max 3 rounds to prevent loops

4. Present Results

Both tools:

## Codex Findings
<findings>

## Gemini Findings
<findings>

## Synthesis
- **Agreements**: Where both models align
- **Divergences**: Where they disagree and your assessment of who's right
- **Action items**: Prioritized changes to consider

Single tool: Present findings with your assessment of which items are valid vs uncertain.

Error Handling

Examples

Code review (both tools):

User: /second-opinions
Agent: [asks tool, task type, focus, diff scope]
User: "Both", "Code review", "Security", "Branch diff"
Agent: [shows diff --stat, dispatches both in parallel]
Agent: [presents synthesis with agreements/divergences]

Plan review (iterative, single tool):

User: /second-opinions review my migration plan with codex
Agent: [asks focus] -> "Correctness & edge cases"
Agent: [extracts plan, dispatches to Codex]
Codex: [returns concerns]
Agent: [revises plan, re-submits via codex-reply]
Codex: [approves revised plan]
Agent: [presents final result]

Custom task:

User: /second-opinions have gemini analyze src/auth/ for race conditions
Agent: [dispatches Gemini with direct file reading]
Agent: [presents findings with assessment]