Analystes en études de marché et spécialistes en marketing Deep research on any topic using Claude, OpenAI, and Gemini in parallel. Dispatches the same research prompt to multiple AI providers, then synthesizes findings into a unified report with cross-source consensus tagging (consensus/majority/contested/unique), citation validation, and a provider comparison matrix.
Activate when user asks to: "research [topic]", "deep research on [topic]", "investigate [topic]", "do a literature review on [topic]", "compare what AI providers say about [topic]", "multi-source analysis of [topic]", or "fact-check [claims] across providers". Also activates for: "research report", "comprehensive report", "cross-validate", "synthesize research".
Do NOT use for: quick factual lookups (use web search instead), single-source questions, code generation, or tasks that don't need multi-provider research.
Workflow: (1) conversational prompt crafting with user approval gate, (2) provider detection + session creation, (3) a self-contained background pipeline run (orchestrate) that dispatches p
2026-06-22