What problem does it solve? Authors holding licensed or owned ranking snapshots lack a safe, auditable way to turn raw leaderboard data into actionable market insight without scraping platforms, storing secrets, or trusting unverifiable LLM output. ## Core Features & Use Cases - Snapshot Import & Analysis: Import authorized local ranking snapshots (permission=owned/licensed) and run decomposition analysis via the Auctra DecompositionProvider with idempotency keys. - LLM-Gated Insights: Reader-demand signals, reusable recipe proposals, and advisory project-fit proposals are produced only from successful, non-simulation LLM decomposition; deterministic code yields observations only. - Auditable Review Flow: Emits redacted receipts (no raw bodies, prompts, or secrets) and routes acceptance exclusively through the existing auctra review accept|partial|reject queue. - Use Case: An author with a licensed Qidian ranking snapshot runs auctra market analyze to receive pending-review recipe proposals and an advisory project-fit assessment, then decides via the standard review queue whether to adopt them into their writing strategy. ## Quick Start Ask the assistant to import your authorized local ranking snapshot with its permission and scope, then run a market analysis with an objective and idempotency key to get reviewable signals and proposals.
