social-intellisted

# social-intel Grounded web-context intelligence on any topic, brand, or account. The skill runs **two independent web searches — Exa (neural, MPP on Tempo) and Tavily (advanced, x402 on Base) via the SELAT Router (x402)** — and the agent fuses them into a single cited brief, cross-checking the two sources and flagging claims only one of them makes. ## When To Use Use when the user wants a grounded, corroborated web read on a topic, brand, or product — not a single-source lookup. The value is in *cross-checking two distinct retrieval methods* (Exa's neural/semantic search vs Tavily's aggregation) so the brief is corroborated rather than dependent on one engine. Every API call is a paid x402 service; the agent does the ranking, sentiment read, and synthesis around the paid data. ## Rails Both steps are native x402, **via the SELAT Router** through the SELAT Router (`rail: via the SELAT Router`): - **x402 on Base**: Exa web search (`api.exa.ai`) — resolves as `x402 on Base`. - **x402 on Base**: Tavily web search (`x402.tavily.com`) — resolves as `x402 on Base`. The `selat` CLI auto-detects each step's protocol at call time. ## Workflow 1. Install: `selat skill install social-intel` 2. Run end-to-end: `selat skill run social-intel --topic "<topic>"` 3. The CLI compiles each step into a `selat-pay` call and prints each result. Recommended agent procedure: 1. **Ground the topic on the web** — Exa `POST /search` (x402 on Base, ~$0.007); returns ranked results with