second-opinion

Second Opinion Skill

Proactively suggest getting another LLM's perspective when the situation warrants it.

Trigger Conditions

Suggest a second opinion when you detect:

1. Architectural Decisions

• Choosing between design patterns (e.g., repository vs service layer)
• Database schema design decisions
• API design choices (REST vs GraphQL, versioning strategy)
• Service decomposition (monolith vs microservices)
• State management approaches

2. Complex Trade-offs

• Performance vs. readability
• Flexibility vs. simplicity
• DRY vs. explicit code
• Build vs. buy decisions
• Consistency vs. availability trade-offs

3. Critical Code Reviews

• Security-sensitive code (authentication, authorization, crypto)
• Performance-critical paths
• Complex algorithms or data structures
• Code handling financial transactions or PII
• Concurrency and threading logic

4. Explicit Requests (trigger words)

• "another perspective"
• "second opinion"
• "sanity check"
• "what do you think"
• "am I on the right track"
• "does this make sense"
• "is this a good approach"

How to Suggest

When conditions are met, offer specific options:

This involves [type of decision]. Would you like a second opinion from another LLM?

• /codex:review - Get OpenAI's analysis via the official Codex Claude Code plugin when installed
• /llm-tools:codex review <scope> - Use gopher-ai's Codex CLI fallback when the official plugin is missing or declined
• /gemini <specific question> - Ask Google Gemini
• /ollama <question> - Use a local model (keeps data private)
• /llm-tools:review-loop --llm fable - Fresh-context Claude subagent review (no external CLI, no extra cost)
• /llm-compare <question> - Compare multiple models

Cross-model rule: a second opinion is most valuable from a different model family than the one that wrote the code. If Claude wrote it, suggest codex/gemini/ollama first. If Codex wrote it (wtcodex flows), suggest the fable review. Never invoke Fable via claude -p — headless print mode bills metered API usage, not the subscription; use the subagent path (or a tmux-driven interactive Claude window when orchestrating from Codex).

Codex routing rule: in Claude Code, prefer /codex:review, /codex:adversarial-review, or /codex:rescue from codex@openai-codex when that official plugin is installed. If it is not installed, use /llm-tools:codex ...; that command warns, prints the official plugin install steps, and can proceed with the existing Codex CLI fallback.

Tailor the suggestion to the context:

For security-sensitive code:

Since this involves authentication logic, you might want a second security review. Try /codex:adversarial-review if the official Codex plugin is installed, /llm-tools:codex review for the CLI fallback, or /ollama (keeps code local) for another perspective.

For architectural decisions:

This is a significant architectural choice. Different models sometimes weigh trade-offs differently. Want to try /llm-compare "should I use X or Y for this use case" to see multiple perspectives?

For complex algorithms:

This algorithm has some complexity. A second set of eyes might catch edge cases. Try /codex:rescue explain the edge cases in this algorithm if the official Codex plugin is installed, or /llm-tools:codex explain the edge cases in this algorithm for the CLI fallback.

When NOT to Suggest

Do not suggest second opinions when:

• User is actively implementing (don't interrupt flow)
• Task is simple/straightforward (typos, formatting, simple fixes)
• User has already made a firm decision
• User said "just do it" or similar
• It's routine code changes with clear requirements
• User previously declined suggestions in this session

Privacy Consideration

Always mention /ollama as an option when the code might be sensitive:

For proprietary code, /ollama keeps everything local - your code never leaves your machine.

Integration with Other Commands

When suggesting, be specific about which command fits best:

Situation	Best Command
Code review	/codex:review when installed; otherwise /llm-tools:codex review
Challenge review	/codex:adversarial-review when installed; otherwise /llm-tools:codex review --ask
Code written by Codex	/llm-tools:review-loop --llm fable (cross-model: Claude reviews Codex's work)
Quick question	/gemini <question>
Sensitive/private code	/ollama <question>
Want multiple views	/llm-compare <question>
Complex reasoning task	/codex:rescue when installed; otherwise /llm-tools:codex or /ollama with larger models