multi-model-triangulation
Made by JeffCross-validate decisions using multiple AI models (Codex, Gemini, Grok). Use when "get a second opinion", evaluating approaches, or high-stakes decisions.
Works With
Quality and Utility come from canonical skill metadata. Community ratings appear separately as stars when subscriber reviews exist.
Subscribe to download
Public previews are available, but downloads, CLI sync, and premium content require an active subscription.
Sign In to Subscribe- Full skill content with deterministic versioning
- CLI sync + local search index
- Access to Jeffrey’s premium library
Save to your account
Saved skills live in your cloud list. Run jsm sync on each machine to pull them locally.
What's New with this Skill
1Updates, enhancements, and release notes for this skill.
New: Multi-Model Triangulation, get a second opinion from Codex, Gemini, or Grok without leaving the chat
Different models have different blind spots. Consensus across two or three models is stronger than any single model's confidence. The skill is the manual relay: Claude generates a copy-paste-ready prompt, you paste it into Codex, Gemini, or Grok, and Claude synthesizes the responses into a unified recommendation.
- Manual relay pattern: Claude generates the prompt, the human pastes it into the other model, the human pastes the response back, Claude synthesizes
- Model strengths table: Claude for nuance and writing, GPT/Codex for implementation details, Gemini for multimodal and recent data, Grok for unconventional takes. Security reviews get all models; for routine code review two is enough
- Synthesis template: consensus (high confidence), divergence table (investigate why models differ), unique insights per model, recommendation, and a confidence level
- Ready-to-copy prompts for the common shapes: idea evaluation (1–10 scoring on quality/utility/feasibility/risk), code review, architecture decision. scripts/format-prompt.py generates them
- Composes with ux-audit (multiple UX perspectives), multi-pass-bug-hunting (cross-validate bug findings), and idea-wizard (score generated ideas)
Claude can't call other models directly, so the skill makes the relay explicit. Claude writes a copy-paste-ready prompt for each model you want to consult, you paste it into Codex, Gemini, or Grok, and you paste the response back. The prompts are specific and structured, and they ask the other model to be critical rather than agree.