Consult external AIs (Gemini 2.5 Pro, OpenAI Codex, Fresh Claude) for second opinions when stuck on bugs or making architectural decisions. Use when debugging attempts have failed, when making significant architectural choices, for security concerns, or when you need a fresh perspective. Automatically suggests consultation after one failed attempt. Provides a synthesis comparing multiple AI perspectives with web-grounded research, a thinking mode, and repo-aware analysis.
Get second opinions from Gemini, OpenAI, or fresh Claude when stuck or making critical decisions.
# Install skill
cd ~/.claude/skills
git clone [this-repo] multi-ai-consultant
# Setup APIs
export GEMINI_API_KEY="your-key" # Required
export OPENAI_API_KEY="sk-..." # Optional
# Copy templates to your project
cp ~/.claude/skills/multi-ai-consultant/templates/GEMINI.md ./
cp ~/.claude/skills/multi-ai-consultant/templates/codex.md ./
cp ~/.claude/skills/multi-ai-consultant/templates/.geminiignore ./
# Start using
# Claude Code will automatically suggest consultations when stuck
This skill activates when you say:
/consult-gemini [question]/consult-codex [question]/consult-claude [question]/consult-ai [question]| AI | Best For | Cost | Speed | |----|----------|------|-------| | Gemini 2.5 Pro | Web research, latest docs, thinking | ~$0.10-0.50 | Slower | | OpenAI Codex | Repo-aware analysis, code review | ~$0.05-0.30 | Medium | | Fresh Claude | Quick opinion, fresh perspective | Free | Fast |
Result: Better solutions, faster debugging, fresh perspectives
You: "Fix this JWT auth bug"
Claude: Tries token expiry fix → Still failing
Claude: "I've tried one approach without success. Should I consult Gemini for a fresh perspective?"
You: "yes"
Claude: Executes /consult-gemini internally
Gemini: Searches web → Finds Cloudflare Workers don't support process.env → Suggests env.JWT_SECRET
Claude: Synthesizes both analyses → Shows 5-part comparison → Asks permission
You: "proceed"
Claude: Implements fix → ✅ Working
Time saved: 20-30 minutes of trial-and-error
# Gemini (required)
npm install -g @google/generative-ai-cli
export GEMINI_API_KEY="your-key"
# Get key: https://aistudio.google.com/apikey
# Codex (optional)
npm install -g codex
export OPENAI_API_KEY="sk-..."
# Get key: https://platform.openai.com/api-keys
cd ~/.claude/skills
git clone [this-repo] multi-ai-consultant
# In your project root
cp ~/.claude/skills/multi-ai-consultant/templates/GEMINI.md ./
cp ~/.claude/skills/multi-ai-consultant/templates/codex.md ./
cp ~/.claude/skills/multi-ai-consultant/templates/.geminiignore ./
# Test Gemini
gemini -p "test" && echo "✅ Working"
# Test Codex (if installed)
codex exec "test" --yolo && echo "✅ Working"
# Ask Claude Code
# "I'm stuck on this bug"
# → Should suggest consultation
Just work normally. Claude Code will suggest consultations when:
# Specific AI
/consult-gemini Is this JWT implementation secure?
/consult-codex Review this codebase for code smells
/consult-claude Quick sanity check on this approach
# Let Claude choose
/consult-ai How should we structure this microservices architecture?
✅ Automatic triggers - Suggests consultation when stuck
✅ Forced synthesis - Prevents parroting external AI
✅ Smart context - Selective file inclusion
✅ Cost tracking - Logs every consultation
✅ Privacy protected - Respects .gitignore + .geminiignore
✅ Three AI options - Gemini (web search), Codex (repo-aware), Fresh Claude (free)
✅ Pre-flight checks - Verifies CLIs work before calling
✅ Error handling - Helpful messages for common issues
Typical consultation:
View spending:
consultation-log-parser.sh --summary
Estimated ROI: 20-30 minutes saved per bug = worth the cost
Automatic protection:
.gitignorenode_modules/, .env*, etc.Extra protection:
.geminiignore for additional exclusionsYou control:
Problem: Claude Code sometimes gets stuck in a mental rut or lacks latest info
Solution: External AIs provide:
Result: Better solutions, faster debugging, informed decisions
❌ Simple bugs you haven't tried fixing yet ❌ Typos or syntax errors ❌ Questions answered by reading docs ❌ Budget is absolute $0 (use Fresh Claude only)
Gemini not working?
npm install -g @google/generative-ai-cli
export GEMINI_API_KEY="your-key"
Codex not working?
npm install -g codex
export OPENAI_API_KEY="sk-..."
Skill not discovered?
ls -la ~/.claude/skills/multi-ai-consultant
# Check SKILL.md exists and has YAML frontmatter
Too expensive?
consultation-log-parser.sh --summarySKILL.mdplanning/multi-ai-consultant-*.mdreferences/examples/Found an issue? Document in SKILL.md Known Issues
Want to add an AI? Create commands/consult-newai.md
Improving synthesis? Edit templates/GEMINI.md or templates/codex.md
MIT
Last Updated: 2025-11-07 Status: Production Ready Maintainer: Jeremy Dawes | jeremy@jezweb.net
Search for places (restaurants, cafes, etc.) via Google Places API proxy on localhost.
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
Create or update AgentSkills. Use when designing, structuring, or packaging skills with scripts, references, and assets.
Start voice calls via the OpenClaw voice-call plugin.
Notion API for creating and managing pages, databases, and blocks.
Gemini CLI for one-shot Q&A, summaries, and generation.
Category:developer