[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6FGP8Z03TKWPRPNSZFN8GKW":3,"skill-same:01M6FGP8Z03TKWPRPNSZFN8GKW":141},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":44,"linkOnly":45,"markdownContent":46,"markdownBody":47,"frontMatter":48,"refName":69,"commitSha":70,"skillDirectory":71,"forksCount":72,"repositoryHtmlUrl":73,"repositoryHomepageUrl":74,"ownerHtmlUrl":75,"ownerBlogUrl":6,"ownerDescription":76,"updatedAt":77,"install":78,"resources":90},"01M6FGP8Z03TKWPRPNSZFN8GKW","multi-ai-consultant",null,"Consult external AIs (Gemini 2.5 Pro, OpenAI Codex, Claude) for second opinions. Use for debugging failures, architectural decisions, security validation, or need fresh perspective with synthesis.","ComeOnOliver\u002Fskillshub","ComeOnOliver","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F43215937?v=4","01M6E5JNFECJAAVFE0W7SMWWD2","Haoran Shu","01M6E5JNNDQS7AJZDYFQ8TPZCK",67,"MIT","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43],"multi ai consultant","multi","consultant","write","consultation","2025","fresh","openai codex","thinking","use","gemini","read","decisions","repo","consult","Bash","openai","debugging","architecture","security","2026-11-10 10:12:30",false,"---\nname: multi-ai-consultant\ndescription: Consult external AIs (Gemini 2.5 Pro, OpenAI Codex, Claude) for second opinions. Use for debugging failures, architectural decisions, security validation, or need fresh perspective with synthesis.\nlicense: MIT\nallowed-tools: [Bash, Read, Task, Write]\nmetadata:\n  version: 1.0.0\n  author: Claude Skills Maintainers\n  last-verified: 2025-11-07\n  production-tested: true\n  keywords:\n    - ai consultation\n    - second opinion\n    - gemini\n    - openai\n    - codex\n    - debugging\n    - architecture\n    - multi-ai\n    - synthesis\n    - fresh perspective\n    - stuck\n    - web research\n    - thinking mode\n    - repo-aware\n---\n\n# Multi-AI Consultant\n\nConsult external AIs for second opinions when Claude Code is stuck or making critical decisions.\n\n---\n\n## What This Skill Does\n\nThis skill enables **future Claude Code sessions** to consult other AIs when:\n- Stuck on a bug after one failed attempt\n- Making architectural decisions\n- Security concerns need validation\n- Fresh perspective needed\n\n**Key innovation**: Uses existing CLI tools (`gemini`, `codex`) instead of building MCP servers - much simpler and more maintainable.\n\n---\n\n## When to Use This Skill\n\n### Automatic Triggers (No User Action Needed)\n\nClaude Code should **automatically suggest** using this skill when:\n\n1. **After 1 failed debugging attempt**\n   - Tried one approach to fix a bug\n   - Still not working or different error\n   - → Suggest: \"Should I consult [Gemini|Fresh Claude] for a second opinion?\"\n\n2. **Before architectural decisions**\n   - Significant design choices (state management, routing, data flow)\n   - Framework selection\n   - Database schema design\n   - → Auto-consult (mention to user): \"Consulting Gemini for architectural validation...\"\n\n3. **Security changes**\n   - Authentication logic\n   - Authorization rules\n   - Cryptography\n   - Input validation\n   - → Auto-consult: \"Consulting Gemini to verify security approach...\"\n\n4. **When uncertain**\n   - Multiple valid approaches\n   - Trade-offs not clear\n   - Conflicting advice in documentation\n   - → Suggest: \"Would you like me to consult another AI for additional perspective?\"\n\n### Manual Invocation (User Commands)\n\nUser can explicitly request consultation with:\n- `\u002Fconsult-gemini [question]` - Gemini 2.5 Pro with thinking, search, grounding\n- `\u002Fconsult-codex [question]` - OpenAI GPT-4 via Codex CLI (repo-aware)\n- `\u002Fconsult-claude [question]` - Fresh Claude subagent (free, fast)\n- `\u002Fconsult-ai [question]` - Router that asks which AI to use\n\n---\n\n## The Three AIs\n\n| AI | Tool | When to Use | Special Features | Cost |\n|----|------|-------------|------------------|------|\n| **Gemini 2.5 Pro** | `gemini` CLI | Web research, latest docs, thinking | Google Search, extended reasoning, grounding | ~$0.10-0.50 |\n| **OpenAI GPT-4** | `codex` CLI | Repo-aware analysis, code review | Auto-scans directory, OpenAI reasoning | ~$0.05-0.30 |\n| **Fresh Claude** | Task tool | Quick second opinion, budget-friendly | Same capabilities, fresh perspective | Free |\n\n**For detailed AI comparison**: Load `references\u002Fai-strengths.md` when choosing which AI to consult for specific use cases.\n\n---\n\n## How It Works\n\n### Architecture\n\n```\nClaude Code encounters bug\u002Fdecision\n        ↓\nSuggests consultation (or user requests)\n        ↓\nUser approves\n        ↓\nExecute appropriate slash command\n        ↓\nCLI command calls external AI\n        ↓\nParse response\n        ↓\nSynthesize: Claude's analysis + External AI's analysis\n        ↓\nPresent 5-part comparison\n        ↓\nAsk permission to implement\n```\n\n### The 5-Part Synthesis Format\n\nEvery consultation must follow this format (prevents parroting):\n\n1. **? My Analysis** - Claude's original reasoning and attempts\n2. **?\u002F?\u002F? Other AI's Analysis** - External AI's complete response\n3. **? Key Differences** - Agreement, divergence, what each AI caught\u002Fmissed\n4. **⚡ Synthesis** - Combined perspective, root cause, trade-offs\n5. **✅ Recommended Action** - Specific next steps with file paths\u002Fline numbers\n\n**End with**: \"Should I proceed with this approach?\"\n\n---\n\n## Setup\n\n**For complete installation guide**: Load `references\u002Fsetup-guide.md` when installing CLIs, configuring API keys, or setting up templates.\n\n**Quick setup**:\n1. **Install CLIs**: `bun add -g @google\u002Fgenerative-ai-cli` (Gemini), `bun add -g codex` (Codex, optional)\n2. **Set API keys**: `export GEMINI_API_KEY=\"...\"`, `export OPENAI_API_KEY=\"...\"`\n3. **Install skill**: Symlink to `~\u002F.claude\u002Fskills\u002Fmulti-ai-consultant`\n4. **Copy templates**: `GEMINI.md`, `codex.md`, `.geminiignore` to project root\n5. **Verify**: `gemini -p \"test\"`, `codex exec - --yolo`\n\nGet API keys:\n- Gemini: https:\u002F\u002Faistudio.google.com\u002Fapikey\n- OpenAI: https:\u002F\u002Fplatform.openai.com\u002Fapi-keys\n\n---\n\n## Usage\n\n**For detailed examples**: Load `references\u002Fusage-examples.md` when learning consultation workflows or seeing real-world scenarios.\n\n**Quick examples**:\n- **Bug**: After 1 failed attempt → `\u002Fconsult-gemini` for web-researched solution\n- **Architecture**: Design decision → `\u002Fconsult-gemini` for latest best practices\n- **Code review**: Refactoring validation → `\u002Fconsult-codex` for repo-aware consistency check\n- **Quick opinion**: Sanity check → `\u002Fconsult-claude` for free fresh perspective\n\n**5 detailed examples available**:\n1. JWT authentication bug (saved ~30 min, found platform-specific issue)\n2. State management choice (informed decision with 2025 patterns)\n3. Refactoring review (found 3 consistency issues)\n4. Security validation (found 2 critical issues via OWASP 2025)\n5. Multi-AI workflow (high-stakes database choice)\n\n---\n\n## Slash Commands\n\n**For complete command reference**: Load `references\u002Fcommands-reference.md` when needing detailed syntax, options, or cost tracking information.\n\n**Quick command overview**:\n\n### \u002Fconsult-gemini [question]\n- **Use**: Web research, latest docs, extended thinking\n- **Features**: Google Search, grounding, thinking mode\n- **Cost**: ~$0.10-0.50\n- **Example**: `\u002Fconsult-gemini Is this JWT secure by 2025 standards?`\n\n### \u002Fconsult-codex [question]\n- **Use**: Repo-aware analysis, code review\n- **Features**: Auto-scans directory, consistency checks\n- **Cost**: ~$0.05-0.30\n- **Example**: `\u002Fconsult-codex Review for performance bottlenecks`\n\n### \u002Fconsult-claude [question]\n- **Use**: Quick second opinion, budget-friendly\n- **Features**: Free, fast, fresh perspective\n- **Cost**: Free\n- **Example**: `\u002Fconsult-claude Am I missing something obvious?`\n\n### \u002Fconsult-ai [question]\n- **Use**: Router (recommends which AI to use)\n- **Features**: Analyzes question, suggests best AI\n- **Cost**: Varies by chosen AI\n- **Example**: `\u002Fconsult-ai How should we structure this architecture?`\n\n---\n\n## Templates\n\nTemplates customize AI behavior for consultations (auto-loaded from project root):\n\n1. **GEMINI.md** - System instructions for Gemini (enforces 5-part format, web search)\n2. **codex.md** - System instructions for Codex (enforces repo-aware analysis)\n3. **.geminiignore** - Privacy exclusions beyond `.gitignore`\n4. **consultation-log-parser.sh** - View consultation history (optional)\n\n**Installation**: Copy from `~\u002F.claude\u002Fskills\u002Fmulti-ai-consultant\u002Ftemplates\u002F` to project root\n\n---\n\n## Privacy & Security\n\n**Automatic protection**:\n- Both CLIs respect `.gitignore` automatically\n- Create `.geminiignore` for extra exclusions (`.env*`, `*secret*`, `*credentials*`)\n- Pre-consultation check warns if sensitive patterns detected\n\n**Privacy best practices**:\n- Always configure `.gitignore` properly\n- Create `.geminiignore` for extra safety\n- Use smart context selection (specific files, not entire repo)\n- Verify what will be sent: `git status --ignored`\n\n**For detailed privacy configuration**: Load `references\u002Fsetup-guide.md` when setting up `.geminiignore` or privacy exclusions.\n\n---\n\n## Cost Tracking\n\nEvery consultation logged to `~\u002F.claude\u002Fai-consultations\u002Fconsultations.log`\n\n**Log format**: `timestamp,ai,model,input_tokens,output_tokens,cost,project_path`\n\n**View logs**: `consultation-log-parser.sh --summary`\n\n**Example output**:\n```\nTotal consultations: 47\nGemini: 23 ($4.25), Codex: 12 ($1.85), Fresh Claude: 12 ($0.00)\nTotal cost: $6.10\n```\n\n**For detailed cost tracking**: Load `references\u002Fcommands-reference.md` when viewing logs, calculating costs, or managing budgets.\n\n---\n\n## Common Issues\n\n**For complete troubleshooting**: Load `references\u002Ftroubleshooting.md` when encountering errors or setup issues.\n\n**Top 5 issues**:\n\n1. **CLI not installed**: `gemini: command not found` → Fix: `bun add -g @google\u002Fgenerative-ai-cli`\n2. **API key invalid**: Authentication failed → Fix: `export GEMINI_API_KEY=\"...\"`\n3. **Context too large**: Token limit exceeded → Fix: Use specific files, not entire repo\n4. **Privacy leak**: `.env` file sent → Fix: Add to `.gitignore` and `.geminiignore`\n5. **Skill not discovered**: Not working → Fix: Check `~\u002F.claude\u002Fskills\u002Fmulti-ai-consultant`\n\n---\n\n## Token Efficiency\n\n### Without This Skill\n\n**Typical scenario** (stuck on bug):\n1. Try approach 1 (~4k tokens)\n2. Research CLI syntax (~3k tokens)\n3. Try approach 2 (~4k tokens)\n4. Research documentation (~3k tokens)\n5. Try approach 3 (~4k tokens)\n\n**Total**: ~20k tokens, 30-45 minutes\n\n### With This Skill\n\n**Same scenario**:\n1. Try approach 1 (~4k tokens)\n2. Execute `\u002Fconsult-gemini` (~1k tokens)\n3. Gemini finds issue (\u003C5k tokens, billed separately)\n4. Implement fix (~3k tokens)\n\n**Total**: ~8k tokens, 5-10 minutes\n\n**Savings**: ~60% tokens, ~75% time\n\n---\n\n## Success Metrics\n\n### Time Efficiency\n- **Without skill**: 30-45 minutes (trial and error)\n- **With skill**: 5-10 minutes (consultation + fix)\n- **Savings**: ~75%\n\n### Token Efficiency\n- **Without skill**: ~20k tokens (multiple attempts)\n- **With skill**: ~8k tokens (one consultation)\n- **Savings**: ~60%\n\n### Error Prevention\n- **Manual CLI use**: 3-5 common errors (flags, parsing, privacy)\n- **With skill**: 0 errors (all handled by commands)\n- **Prevention**: 100%\n\n### Quality\n- **Manual**: Risk of not synthesizing (just copying external AI)\n- **With skill**: Forced synthesis via `GEMINI.md`\u002F`codex.md`\n- **Improvement**: Guaranteed value-add\n\n---\n\n## Why CLI Approach (Not MCP)?\n\n| Aspect | MCP Server | CLI Approach |\n|--------|------------|--------------|\n| Setup time | 4-6 hours | 60-75 minutes |\n| Complexity | High (MCP protocol) | Low (bash + CLIs) |\n| Maintenance | Update MCP SDK | Update CLI (rare) |\n| Flexibility | Locked to AIs | Any AI with CLI |\n| Debugging | MCP protocol | Standard bash |\n| Dependencies | MCP SDK, npm | Just CLIs |\n\n**Winner**: CLI approach - 80% less effort, same functionality\n\n---\n\n## When to Load References\n\nLoad reference files when working on specific aspects of AI consultation:\n\n### ai-strengths.md\nLoad when:\n- **Selection-based**: Choosing which AI to consult (Gemini vs Codex vs Fresh Claude)\n- **Comparison-based**: Understanding capabilities, costs, and trade-offs between AIs\n- **Strategy-based**: Planning combination strategies (free → paid, paid first, budget-conscious)\n- **Capability-based**: Understanding special features (Google Search, extended thinking, grounding, repo-aware, fresh perspective)\n- **Scenario-based**: Multiple AI consultation workflows, validation workflows\n\n### setup-guide.md\nLoad when:\n- **Installation-based**: Setting up Gemini CLI, Codex CLI, or installing the skill\n- **Configuration-based**: Configuring API keys, environment variables, or system paths\n- **Privacy-based**: Setting up .geminiignore, privacy exclusions, or security configurations\n- **Template-based**: Installing GEMINI.md, codex.md, .geminiignore, or consultation-log-parser.sh\n- **Verification-based**: Testing CLI installation, API keys, or skill discovery\n\n### commands-reference.md\nLoad when:\n- **Command-based**: Using \u002Fconsult-gemini, \u002Fconsult-codex, \u002Fconsult-claude, or \u002Fconsult-ai\n- **Syntax-based**: Understanding command flags, options, or context selection\n- **Cost-based**: Understanding cost tracking, log format, or viewing consultation history\n- **Logging-based**: Using consultation-log-parser.sh or analyzing consultation patterns\n\n### usage-examples.md\nLoad when:\n- **Scenario-based**: Learning how to use skill for bugs, architecture decisions, or code review\n- **Workflow-based**: Understanding consultation workflow, synthesis process, or multi-AI approach\n- **Example-based**: Seeing real-world examples of consultations and their outcomes (5 detailed examples available)\n\n### troubleshooting.md\nLoad when:\n- **Error-based**: Encountering specific errors (command not found, API key invalid, parsing failures)\n- **Diagnosis-based**: Troubleshooting CLI issues, API connectivity, or skill discovery\n- **Fix-based**: Resolving known issues with step-by-step solutions (8 common issues documented)\n- **Debugging-based**: Testing CLIs manually, checking configurations, or verifying installations\n\n---\n\n## Contributing\n\n**Found an issue?**\n- Document it in troubleshooting.md\n- Include fix\u002Fworkaround\n- Update slash commands to prevent\n\n**Adding new AI?**\n- Create new slash command: `commands\u002Fconsult-newai.md`\n- Add to router: Update `commands\u002Fconsult-ai.md`\n- Create template: `templates\u002Fnewai.md` (if CLI supports system instructions)\n- Update documentation\n\n**Improving synthesis?**\n- Edit templates: `templates\u002FGEMINI.md`, `templates\u002Fcodex.md`\n- Test with real consultations\n- Measure before\u002Fafter quality\n\n---\n\n## References\n\n### External Resources\n\n- **Gemini CLI**: https:\u002F\u002Fai.google.dev\u002Fgemini-api\u002Fdocs\u002Fcli\n- **OpenAI Codex**: https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002Fcodex\n- **OpenAI API**: https:\u002F\u002Fplatform.openai.com\u002Fdocs\n- **Gemini API Pricing**: https:\u002F\u002Fai.google.dev\u002Fpricing\n- **OpenAI Pricing**: https:\u002F\u002Fopenai.com\u002Fpricing\n\n### Internal Files\n\n- **Planning docs**: `planning\u002Fmulti-ai-consultant-*.md`\n- **Slash commands**: `commands\u002F*.md`\n- **Templates**: `templates\u002F*`\n- **Scripts**: `scripts\u002F*`\n- **References**: `references\u002F*.md` (5 reference files)\n\n---\n\n## License\n\nMIT License - See LICENSE file\n\n---\n\n**Last Updated**: 2025-11-07\n**Status**: Production Ready\n**Maintainer**: Claude Skills Maintainers | maintainers@example.com\n","# Multi-AI Consultant\n\nConsult external AIs for second opinions when Claude Code is stuck or making critical decisions.\n\n---\n\n## What This Skill Does\n\nThis skill enables **future Claude Code sessions** to consult other AIs when:\n- Stuck on a bug after one failed attempt\n- Making architectural decisions\n- Security concerns need validation\n- Fresh perspective needed\n\n**Key innovation**: Uses existing CLI tools (`gemini`, `codex`) instead of building MCP servers - much simpler and more maintainable.\n\n---\n\n## When to Use This Skill\n\n### Automatic Triggers (No User Action Needed)\n\nClaude Code should **automatically suggest** using this skill when:\n\n1. **After 1 failed debugging attempt**\n   - Tried one approach to fix a bug\n   - Still not working or different error\n   - → Suggest: \"Should I consult [Gemini|Fresh Claude] for a second opinion?\"\n\n2. **Before architectural decisions**\n   - Significant design choices (state management, routing, data flow)\n   - Framework selection\n   - Database schema design\n   - → Auto-consult (mention to user): \"Consulting Gemini for architectural validation...\"\n\n3. **Security changes**\n   - Authentication logic\n   - Authorization rules\n   - Cryptography\n   - Input validation\n   - → Auto-consult: \"Consulting Gemini to verify security approach...\"\n\n4. **When uncertain**\n   - Multiple valid approaches\n   - Trade-offs not clear\n   - Conflicting advice in documentation\n   - → Suggest: \"Would you like me to consult another AI for additional perspective?\"\n\n### Manual Invocation (User Commands)\n\nUser can explicitly request consultation with:\n- `\u002Fconsult-gemini [question]` - Gemini 2.5 Pro with thinking, search, grounding\n- `\u002Fconsult-codex [question]` - OpenAI GPT-4 via Codex CLI (repo-aware)\n- `\u002Fconsult-claude [question]` - Fresh Claude subagent (free, fast)\n- `\u002Fconsult-ai [question]` - Router that asks which AI to use\n\n---\n\n## The Three AIs\n\n| AI | Tool | When to Use | Special Features | Cost |\n|----|------|-------------|------------------|------|\n| **Gemini 2.5 Pro** | `gemini` CLI | Web research, latest docs, thinking | Google Search, extended reasoning, grounding | ~$0.10-0.50 |\n| **OpenAI GPT-4** | `codex` CLI | Repo-aware analysis, code review | Auto-scans directory, OpenAI reasoning | ~$0.05-0.30 |\n| **Fresh Claude** | Task tool | Quick second opinion, budget-friendly | Same capabilities, fresh perspective | Free |\n\n**For detailed AI comparison**: Load `references\u002Fai-strengths.md` when choosing which AI to consult for specific use cases.\n\n---\n\n## How It Works\n\n### Architecture\n\n```\nClaude Code encounters bug\u002Fdecision\n        ↓\nSuggests consultation (or user requests)\n        ↓\nUser approves\n        ↓\nExecute appropriate slash command\n        ↓\nCLI command calls external AI\n        ↓\nParse response\n        ↓\nSynthesize: Claude's analysis + External AI's analysis\n        ↓\nPresent 5-part comparison\n        ↓\nAsk permission to implement\n```\n\n### The 5-Part Synthesis Format\n\nEvery consultation must follow this format (prevents parroting):\n\n1. **? My Analysis** - Claude's original reasoning and attempts\n2. **?\u002F?\u002F? Other AI's Analysis** - External AI's complete response\n3. **? Key Differences** - Agreement, divergence, what each AI caught\u002Fmissed\n4. **⚡ Synthesis** - Combined perspective, root cause, trade-offs\n5. **✅ Recommended Action** - Specific next steps with file paths\u002Fline numbers\n\n**End with**: \"Should I proceed with this approach?\"\n\n---\n\n## Setup\n\n**For complete installation guide**: Load `references\u002Fsetup-guide.md` when installing CLIs, configuring API keys, or setting up templates.\n\n**Quick setup**:\n1. **Install CLIs**: `bun add -g @google\u002Fgenerative-ai-cli` (Gemini), `bun add -g codex` (Codex, optional)\n2. **Set API keys**: `export GEMINI_API_KEY=\"...\"`, `export OPENAI_API_KEY=\"...\"`\n3. **Install skill**: Symlink to `~\u002F.claude\u002Fskills\u002Fmulti-ai-consultant`\n4. **Copy templates**: `GEMINI.md`, `codex.md`, `.geminiignore` to project root\n5. **Verify**: `gemini -p \"test\"`, `codex exec - --yolo`\n\nGet API keys:\n- Gemini: https:\u002F\u002Faistudio.google.com\u002Fapikey\n- OpenAI: https:\u002F\u002Fplatform.openai.com\u002Fapi-keys\n\n---\n\n## Usage\n\n**For detailed examples**: Load `references\u002Fusage-examples.md` when learning consultation workflows or seeing real-world scenarios.\n\n**Quick examples**:\n- **Bug**: After 1 failed attempt → `\u002Fconsult-gemini` for web-researched solution\n- **Architecture**: Design decision → `\u002Fconsult-gemini` for latest best practices\n- **Code review**: Refactoring validation → `\u002Fconsult-codex` for repo-aware consistency check\n- **Quick opinion**: Sanity check → `\u002Fconsult-claude` for free fresh perspective\n\n**5 detailed examples available**:\n1. JWT authentication bug (saved ~30 min, found platform-specific issue)\n2. State management choice (informed decision with 2025 patterns)\n3. Refactoring review (found 3 consistency issues)\n4. Security validation (found 2 critical issues via OWASP 2025)\n5. Multi-AI workflow (high-stakes database choice)\n\n---\n\n## Slash Commands\n\n**For complete command reference**: Load `references\u002Fcommands-reference.md` when needing detailed syntax, options, or cost tracking information.\n\n**Quick command overview**:\n\n### \u002Fconsult-gemini [question]\n- **Use**: Web research, latest docs, extended thinking\n- **Features**: Google Search, grounding, thinking mode\n- **Cost**: ~$0.10-0.50\n- **Example**: `\u002Fconsult-gemini Is this JWT secure by 2025 standards?`\n\n### \u002Fconsult-codex [question]\n- **Use**: Repo-aware analysis, code review\n- **Features**: Auto-scans directory, consistency checks\n- **Cost**: ~$0.05-0.30\n- **Example**: `\u002Fconsult-codex Review for performance bottlenecks`\n\n### \u002Fconsult-claude [question]\n- **Use**: Quick second opinion, budget-friendly\n- **Features**: Free, fast, fresh perspective\n- **Cost**: Free\n- **Example**: `\u002Fconsult-claude Am I missing something obvious?`\n\n### \u002Fconsult-ai [question]\n- **Use**: Router (recommends which AI to use)\n- **Features**: Analyzes question, suggests best AI\n- **Cost**: Varies by chosen AI\n- **Example**: `\u002Fconsult-ai How should we structure this architecture?`\n\n---\n\n## Templates\n\nTemplates customize AI behavior for consultations (auto-loaded from project root):\n\n1. **GEMINI.md** - System instructions for Gemini (enforces 5-part format, web search)\n2. **codex.md** - System instructions for Codex (enforces repo-aware analysis)\n3. **.geminiignore** - Privacy exclusions beyond `.gitignore`\n4. **consultation-log-parser.sh** - View consultation history (optional)\n\n**Installation**: Copy from `~\u002F.claude\u002Fskills\u002Fmulti-ai-consultant\u002Ftemplates\u002F` to project root\n\n---\n\n## Privacy & Security\n\n**Automatic protection**:\n- Both CLIs respect `.gitignore` automatically\n- Create `.geminiignore` for extra exclusions (`.env*`, `*secret*`, `*credentials*`)\n- Pre-consultation check warns if sensitive patterns detected\n\n**Privacy best practices**:\n- Always configure `.gitignore` properly\n- Create `.geminiignore` for extra safety\n- Use smart context selection (specific files, not entire repo)\n- Verify what will be sent: `git status --ignored`\n\n**For detailed privacy configuration**: Load `references\u002Fsetup-guide.md` when setting up `.geminiignore` or privacy exclusions.\n\n---\n\n## Cost Tracking\n\nEvery consultation logged to `~\u002F.claude\u002Fai-consultations\u002Fconsultations.log`\n\n**Log format**: `timestamp,ai,model,input_tokens,output_tokens,cost,project_path`\n\n**View logs**: `consultation-log-parser.sh --summary`\n\n**Example output**:\n```\nTotal consultations: 47\nGemini: 23 ($4.25), Codex: 12 ($1.85), Fresh Claude: 12 ($0.00)\nTotal cost: $6.10\n```\n\n**For detailed cost tracking**: Load `references\u002Fcommands-reference.md` when viewing logs, calculating costs, or managing budgets.\n\n---\n\n## Common Issues\n\n**For complete troubleshooting**: Load `references\u002Ftroubleshooting.md` when encountering errors or setup issues.\n\n**Top 5 issues**:\n\n1. **CLI not installed**: `gemini: command not found` → Fix: `bun add -g @google\u002Fgenerative-ai-cli`\n2. **API key invalid**: Authentication failed → Fix: `export GEMINI_API_KEY=\"...\"`\n3. **Context too large**: Token limit exceeded → Fix: Use specific files, not entire repo\n4. **Privacy leak**: `.env` file sent → Fix: Add to `.gitignore` and `.geminiignore`\n5. **Skill not discovered**: Not working → Fix: Check `~\u002F.claude\u002Fskills\u002Fmulti-ai-consultant`\n\n---\n\n## Token Efficiency\n\n### Without This Skill\n\n**Typical scenario** (stuck on bug):\n1. Try approach 1 (~4k tokens)\n2. Research CLI syntax (~3k tokens)\n3. Try approach 2 (~4k tokens)\n4. Research documentation (~3k tokens)\n5. Try approach 3 (~4k tokens)\n\n**Total**: ~20k tokens, 30-45 minutes\n\n### With This Skill\n\n**Same scenario**:\n1. Try approach 1 (~4k tokens)\n2. Execute `\u002Fconsult-gemini` (~1k tokens)\n3. Gemini finds issue (\u003C5k tokens, billed separately)\n4. Implement fix (~3k tokens)\n\n**Total**: ~8k tokens, 5-10 minutes\n\n**Savings**: ~60% tokens, ~75% time\n\n---\n\n## Success Metrics\n\n### Time Efficiency\n- **Without skill**: 30-45 minutes (trial and error)\n- **With skill**: 5-10 minutes (consultation + fix)\n- **Savings**: ~75%\n\n### Token Efficiency\n- **Without skill**: ~20k tokens (multiple attempts)\n- **With skill**: ~8k tokens (one consultation)\n- **Savings**: ~60%\n\n### Error Prevention\n- **Manual CLI use**: 3-5 common errors (flags, parsing, privacy)\n- **With skill**: 0 errors (all handled by commands)\n- **Prevention**: 100%\n\n### Quality\n- **Manual**: Risk of not synthesizing (just copying external AI)\n- **With skill**: Forced synthesis via `GEMINI.md`\u002F`codex.md`\n- **Improvement**: Guaranteed value-add\n\n---\n\n## Why CLI Approach (Not MCP)?\n\n| Aspect | MCP Server | CLI Approach |\n|--------|------------|--------------|\n| Setup time | 4-6 hours | 60-75 minutes |\n| Complexity | High (MCP protocol) | Low (bash + CLIs) |\n| Maintenance | Update MCP SDK | Update CLI (rare) |\n| Flexibility | Locked to AIs | Any AI with CLI |\n| Debugging | MCP protocol | Standard bash |\n| Dependencies | MCP SDK, npm | Just CLIs |\n\n**Winner**: CLI approach - 80% less effort, same functionality\n\n---\n\n## When to Load References\n\nLoad reference files when working on specific aspects of AI consultation:\n\n### ai-strengths.md\nLoad when:\n- **Selection-based**: Choosing which AI to consult (Gemini vs Codex vs Fresh Claude)\n- **Comparison-based**: Understanding capabilities, costs, and trade-offs between AIs\n- **Strategy-based**: Planning combination strategies (free → paid, paid first, budget-conscious)\n- **Capability-based**: Understanding special features (Google Search, extended thinking, grounding, repo-aware, fresh perspective)\n- **Scenario-based**: Multiple AI consultation workflows, validation workflows\n\n### setup-guide.md\nLoad when:\n- **Installation-based**: Setting up Gemini CLI, Codex CLI, or installing the skill\n- **Configuration-based**: Configuring API keys, environment variables, or system paths\n- **Privacy-based**: Setting up .geminiignore, privacy exclusions, or security configurations\n- **Template-based**: Installing GEMINI.md, codex.md, .geminiignore, or consultation-log-parser.sh\n- **Verification-based**: Testing CLI installation, API keys, or skill discovery\n\n### commands-reference.md\nLoad when:\n- **Command-based**: Using \u002Fconsult-gemini, \u002Fconsult-codex, \u002Fconsult-claude, or \u002Fconsult-ai\n- **Syntax-based**: Understanding command flags, options, or context selection\n- **Cost-based**: Understanding cost tracking, log format, or viewing consultation history\n- **Logging-based**: Using consultation-log-parser.sh or analyzing consultation patterns\n\n### usage-examples.md\nLoad when:\n- **Scenario-based**: Learning how to use skill for bugs, architecture decisions, or code review\n- **Workflow-based**: Understanding consultation workflow, synthesis process, or multi-AI approach\n- **Example-based**: Seeing real-world examples of consultations and their outcomes (5 detailed examples available)\n\n### troubleshooting.md\nLoad when:\n- **Error-based**: Encountering specific errors (command not found, API key invalid, parsing failures)\n- **Diagnosis-based**: Troubleshooting CLI issues, API connectivity, or skill discovery\n- **Fix-based**: Resolving known issues with step-by-step solutions (8 common issues documented)\n- **Debugging-based**: Testing CLIs manually, checking configurations, or verifying installations\n\n---\n\n## Contributing\n\n**Found an issue?**\n- Document it in troubleshooting.md\n- Include fix\u002Fworkaround\n- Update slash commands to prevent\n\n**Adding new AI?**\n- Create new slash command: `commands\u002Fconsult-newai.md`\n- Add to router: Update `commands\u002Fconsult-ai.md`\n- Create template: `templates\u002Fnewai.md` (if CLI supports system instructions)\n- Update documentation\n\n**Improving synthesis?**\n- Edit templates: `templates\u002FGEMINI.md`, `templates\u002Fcodex.md`\n- Test with real consultations\n- Measure before\u002Fafter quality\n\n---\n\n## References\n\n### External Resources\n\n- **Gemini CLI**: https:\u002F\u002Fai.google.dev\u002Fgemini-api\u002Fdocs\u002Fcli\n- **OpenAI Codex**: https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002Fcodex\n- **OpenAI API**: https:\u002F\u002Fplatform.openai.com\u002Fdocs\n- **Gemini API Pricing**: https:\u002F\u002Fai.google.dev\u002Fpricing\n- **OpenAI Pricing**: https:\u002F\u002Fopenai.com\u002Fpricing\n\n### Internal Files\n\n- **Planning docs**: `planning\u002Fmulti-ai-consultant-*.md`\n- **Slash commands**: `commands\u002F*.md`\n- **Templates**: `templates\u002F*`\n- **Scripts**: `scripts\u002F*`\n- **References**: `references\u002F*.md` (5 reference files)\n\n---\n\n## License\n\nMIT License - See LICENSE file\n\n---\n\n**Last Updated**: 2025-11-07\n**Status**: Production Ready\n**Maintainer**: Claude Skills Maintainers | maintainers@example.com\n",{"name":5,"license":15,"metadata":49,"description":7,"allowed-tools":65},{"author":50,"version":51,"keywords":52,"last-verified":63,"production-tested":64},"Claude Skills Maintainers","1.0.0",[53,54,34,40,55,41,42,56,57,58,59,60,61,62],"ai consultation","second opinion","codex","multi-ai","synthesis","fresh perspective","stuck","web research","thinking mode","repo-aware",1762473600000,true,[39,66,67,68],"Read","Task","Write","main","def8531e65114c0fca8fb8551c1871ee0eed705c","skills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant",24,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub","https:\u002F\u002Fskillshub.wtf","https:\u002F\u002Fgithub.com\u002FComeOnOliver","AI & Blockchain Engineer\r\n","2026-06-24 22:07:19",{"source":8,"skillName":5,"commands":79,"prompt":89,"zipAvailable":64},[80,83,86],{"tool":81,"command":82},"npx","npx skills add ComeOnOliver\u002Fskillshub --skill multi-ai-consultant",{"tool":84,"command":85},"bunx","bunx skills add ComeOnOliver\u002Fskillshub --skill multi-ai-consultant",{"tool":87,"command":88},"pnpm","pnpm dlx skills add ComeOnOliver\u002Fskillshub --skill multi-ai-consultant","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002FSKILL.md",[91,97,101,105,109,113,117,123,129,133,137],{"path":92,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":95,"sourceHtmlUrl":96},"examples\u002Fsample-consultation.md","REFERENCE","参考文档",9800,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Fexamples\u002Fsample-consultation.md",{"path":98,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":99,"sourceHtmlUrl":100},"references\u002Fai-strengths.md",6170,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Freferences\u002Fai-strengths.md",{"path":102,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":103,"sourceHtmlUrl":104},"references\u002Fcommands-reference.md",13155,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Freferences\u002Fcommands-reference.md",{"path":106,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":107,"sourceHtmlUrl":108},"references\u002Fsetup-guide.md",10235,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Freferences\u002Fsetup-guide.md",{"path":110,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":111,"sourceHtmlUrl":112},"references\u002Ftroubleshooting.md",14130,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Freferences\u002Ftroubleshooting.md",{"path":114,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":115,"sourceHtmlUrl":116},"references\u002Fusage-examples.md",15488,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Freferences\u002Fusage-examples.md",{"path":118,"resourceType":119,"resourceTypeDesc":120,"sizeBytes":121,"sourceHtmlUrl":122},"scripts\u002Fsetup-apis.sh","SCRIPT","脚本",5593,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Fscripts\u002Fsetup-apis.sh",{"path":124,"resourceType":125,"resourceTypeDesc":126,"sizeBytes":127,"sourceHtmlUrl":128},"templates\u002F.geminiignore","OTHER","其他",3052,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Ftemplates\u002F.geminiignore",{"path":130,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":131,"sourceHtmlUrl":132},"templates\u002Fcodex.md",7870,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Ftemplates\u002Fcodex.md",{"path":134,"resourceType":119,"resourceTypeDesc":120,"sizeBytes":135,"sourceHtmlUrl":136},"templates\u002Fconsultation-log-parser.sh",4909,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Ftemplates\u002Fconsultation-log-parser.sh",{"path":138,"resourceType":93,"resourceTypeDesc":94,"sizeBytes":139,"sourceHtmlUrl":140},"templates\u002FGEMINI.md",7288,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fsecondsky\u002Fclaude-skills\u002Fmulti-ai-consultant\u002Ftemplates\u002FGEMINI.md",[142,152,162,171,177,184],{"skillNo":143,"name":144,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":145,"categories":146,"keywords":148,"publishedAt":44},"01M6FKJN13GMG4KY57J3QD8GY5","model-hierarchy-skill","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzscole\u002Fmodel-hierarchy-skill\u002Fmodel-hierarchy-skill\u002FSKILL.md",[147],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[149,150,151],"model","model hierarchy","model hierarchy skill",{"skillNo":153,"name":154,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":155,"categories":156,"keywords":158,"publishedAt":44},"01M6FKJJJ6GCAMR4PPPRD12M5X","zoom-mcp","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fzoom-mcp\u002FSKILL.md",[157],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[159,160,161],"MCP","zoom","zoom mcp",{"skillNo":163,"name":164,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":165,"categories":166,"keywords":167,"publishedAt":44},"01M6FKJE3KYT3KPEZMEEHYKA6K","zoom-apps-sdk","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fzoom-apps-sdk\u002FSKILL.md",[],[168,169,170,160],"sdk","zoom apps sdk","apps",{"skillNo":172,"name":173,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":174,"categories":175,"keywords":176,"publishedAt":44},"01M6FKJ9EDH9Q5VWDF8HAJRSP0","windows","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwindows\u002FSKILL.md",[],[173],{"skillNo":178,"name":179,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":180,"categories":181,"keywords":183,"publishedAt":44},"01M6FKJ666BEEFC3AE9GJ72Q93","whiteboard","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwhiteboard\u002FSKILL.md",[182],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[179],{"skillNo":185,"name":186,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":187,"categories":188,"keywords":189,"publishedAt":44},"01M6FKJ21WFCDGTJQ2Q91T9DVW","websockets","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwebsockets\u002FSKILL.md",[],[186]]