second-opinion
This skill queries GPT-5.2 Pro and Gemini 3 Pro in parallel to get alternative perspectives on Claude's outputs. It extracts context automatically from conversation history, sends the same task to both models concurrently via subagents, then presents all responses for Claude to perform objective comparative analysis. The skill should be used when the user types `/second-opinion` to refine or validate Claude's output across any content type: answers, drafts, code, strategies, creative work, or analysis.
Best use case
second-opinion is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
This skill queries GPT-5.2 Pro and Gemini 3 Pro in parallel to get alternative perspectives on Claude's outputs. It extracts context automatically from conversation history, sends the same task to both models concurrently via subagents, then presents all responses for Claude to perform objective comparative analysis. The skill should be used when the user types `/second-opinion` to refine or validate Claude's output across any content type: answers, drafts, code, strategies, creative work, or analysis.
Teams using second-opinion should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/second-opinion/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How second-opinion Compares
| Feature / Agent | second-opinion | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
This skill queries GPT-5.2 Pro and Gemini 3 Pro in parallel to get alternative perspectives on Claude's outputs. It extracts context automatically from conversation history, sends the same task to both models concurrently via subagents, then presents all responses for Claude to perform objective comparative analysis. The skill should be used when the user types `/second-opinion` to refine or validate Claude's output across any content type: answers, drafts, code, strategies, creative work, or analysis.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Second Opinion
## Overview
This skill enables Claude to get second (and third) opinions on its outputs by querying GPT-5.2 Pro and Gemini 3 Pro in parallel. It automatically extracts the original task and Claude's response from conversation history, submits the same prompt to both alternative models via concurrent async API calls, then presents all three responses for Claude to analyze objectively.
**Claude performs the actual comparative analysis** - the skill just fetches and formats the responses.
## When to Use This Skill
Invoke this skill using `/second-opinion` after Claude has provided a response and you want:
- Validation of accuracy or completeness
- Alternative approaches or perspectives
- Comparison across multiple models
- An objective determination of which answer is best
## Workflow
```
/second-opinion → Validate API Keys
→ Extract Context (auto)
→ Query ChatGPT + Gemini (parallel via asyncio)
→ Format All Responses
→ Claude Analyzes & Compares
```
## Setup Requirements
### API Keys
Both API keys must be set as environment variables:
```bash
export OPENAI_API_KEY="your-openai-key-here"
export GEMINI_API_KEY="your-google-ai-key-here"
```
**Where to get keys:**
- **OpenAI Key**: https://platform.openai.com/api-keys
- **Gemini Key**: https://aistudio.google.com/apikey
### Dependencies
```bash
pip install aiohttp
```
## How It Works
### 1. Context Extraction
When `/second-opinion` is invoked, the skill automatically:
- Parses the conversation history
- Identifies the original user task
- Extracts Claude's current response
### 2. Parallel Model Querying
Both models are queried simultaneously using `asyncio.gather()`:
- GPT-5.2 Pro via OpenAI API
- Gemini 3 Pro via Google Generative AI API
Both receive identical prompts for fair comparison. Includes retry logic with exponential backoff.
### 3. Response Presentation
The skill formats all three responses in markdown:
```markdown
# Second Opinion Results
## Original Task
[The user's question/task]
---
## Claude's Response
[Claude's original response]
---
## GPT-5.2 Pro's Response
[ChatGPT's response]
---
## Gemini 3 Pro's Response
[Gemini's response]
---
*Response times: ChatGPT: 2.3s | Gemini: 1.8s*
```
### 4. Claude's Analysis
Claude then provides an impartial analysis:
- **Accuracy**: Which responses are factually correct?
- **Completeness**: Which best addresses all aspects?
- **Clarity**: Which is most clear and organized?
- **Recommendation**: Which response is best and why?
Claude is explicitly prompted to be impartial and acknowledge if its original response was not the best.
## Error Handling
- **Partial failure**: If one model fails, the comparison continues with available responses
- **Total failure**: Clear error message with troubleshooting guidance
## File Structure
```
second-opinion/
├── SKILL.md
└── scripts/
├── main.py # Async orchestrator
├── api_clients.py # Async OpenAI + Gemini clients
├── extract_context.py # Context extraction
├── analyze_responses.py # Basic utilities
└── format_output.py # Response formatting
```
## How to Invoke
The script accepts JSON input via stdin. Pipe a JSON object with:
- `conversation_history`: List of message objects with `role` and `content`
- `claude_response`: Claude's response string
Example invocation:
```bash
cd ~/.claude/skills/second-opinion/scripts && echo '{"conversation_history": [{"role": "user", "content": "What is X?"}], "claude_response": "X is..."}' | python main.py
```
## Technical Details
- **Parallel Execution**: Uses `asyncio.gather()` for true concurrent I/O
- **Timeout**: 300 seconds (5 minutes) per model
- **Retry Logic**: 2 retries with exponential backoff
- **Models**: `gpt-5.2-pro` and `gemini-3-pro`
## Limitations
- Both API keys must be configured
- Works best when Claude has already provided a complete response
- Requires `aiohttp` packageRelated Skills
second-brain
Personal intelligence system for capturing thoughts, managing knowledge, and surfacing insights. Use when user wants to capture an idea, task, or note during conversation; query their knowledge base; check their inbox; review digests; or update task status. Triggers include "remember this," "add a task," "what did I say about," "show my inbox," or "mark complete."
bgo
Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.
synthflow-ai-automation
Automate Synthflow AI tasks via Rube MCP (Composio). Always search tools first for current schemas.
sync-agents
Synchronize GitHub Copilot instructions, custom agents, and skills into detected AI coding agent configurations in this repository. Use when asked to mirror .github/copilot-instructions.md, .github/instructions, .github/agents, or .github/skills into Claude, Codex, Cursor, Gemini, Windsurf, and related tooling.
synapse
Multi-AI Agent Orchestration System with configurable models and role-based workflows. Use when you need to coordinate multiple AI agents (Claude, Gemini, Codex) for complex tasks like planning, code generation, analysis, review, and execution. Supports agentic workflow patterns: parallel specialists, pipeline, and swarm orchestration. Compatible with Claude Code, Cursor, and OpenCode. Triggers: 'orchestrate agents', 'multi-agent workflow', 'plan and execute', 'code review pipeline', 'run synapse', 'agentic workflow'.
synapse-action-development
Explains how to create Synapse plugin actions. Use when the user asks to "create an action", "write an action", uses "@action decorator", "BaseAction class", "function-based action", "class-based action", "Pydantic params", "ActionPipeline", "DataType", "input_type", "output_type", "semantic types", "YOLODataset", "ModelWeights", "pipeline chaining", or needs help with synapse plugin action development.
swot-pestle-analysis
Strategic environmental analysis using SWOT, PESTLE, and Porter's Five Forces. Creates structured assessments with Mermaid visualizations for competitive positioning and strategic planning.
switchailocal
Unified LLM proxy for AI agents. Route all model requests through http://localhost:18080/v1. Provides FREE access to Gemini CLI, Claude CLI, Codex, and Vibe via your existing subscriptions. Use when: (1) making LLM calls using provider prefixes, (2) switching between CLI/Local/Cloud providers, (3) needing to attach local files/folders to prompts via CLI, (4) requiring intelligent routing between models, or (5) needing to monitor provider health and analytics.
swift-actor-persistence
Use when building a thread-safe data persistence layer in Swift using actors with in-memory cache and file storage.
swamp-vault
Manage swamp vaults for secure secret storage. Use when creating vaults, storing secrets, retrieving secrets, listing vault keys, or working with vault expressions in workflows. Triggers on "vault", "secret", "secrets", "credentials", "api key storage", "secure storage", "password", "token", "key management", "sensitive data", "encrypt", "aws secrets manager", "store secret", "put secret", "get secret", "credential storage", or vault-related CLI commands.
supadata-automation
Automate Supadata tasks via Rube MCP (Composio). Always search tools first for current schemas.
Suno Song Creator
This skill should be used when the user asks to "create a Suno prompt", "write a Suno song", "generate music with Suno", "help me with Suno", "make a song prompt", "create lyrics for Suno", "build a music prompt", or mentions Suno AI music generation. Provides comprehensive guidance for creating professional Suno prompts using advanced prompting strategies, structured formatting within 1000 character limit (NO blank lines between sections), parameter optimization, genre-specific techniques, interactive questioning with efficient project name collection, automated artist/song research via sub-agent (web fetching + pattern extraction), automatic file export to organized project directories, AI-slop avoidance for authentic human-centered lyrics, copyright-safe style descriptions that avoid artist/album/song names, character counting utilities for accurate verification, and optional independent quality review via sub-agent for professional assessment.