Search Discovery Relevance Review

Use this skill to convert a search, discovery, relevance quality, filters, ranking drift, zero-results, and user trust question into a concrete artifact with owners, gates, metrics, and recovery paths.

Workflow

1. Identify searchable corpus, user intents, query types, ranking surfaces, business incentives, cold-start items, and harm from bad relevance.
2. Read references/search-discovery-relevance-patterns.md.
3. Separate retrieval, filtering, ranking, personalization, editorial curation, sponsored placement, and moderation eligibility.
4. Define quality evals, query logs, zero-result recovery, feedback signals, fairness checks, ranking-change rollout, and support/debug tools.
5. Produce relevance audit, state machine, decision table, event schema, eval set, and launch checklist.

When not to use

• Do not use for generic advice the base model already handles without this skill's specific artifact contract.

Guardrails