strategic-search

Constraint-based problem solving automates structured web searches across multiple engines to gather, rank, and validate candidate solutions against defined criteria. It extracts constraints from your problem, ranks them by uniqueness, and validates candidates.

Multi-engine search orchestration defaults to DuckDuckGo while utilizing Brave and Serper for cross-validation and deeper exploration. This ensures candidate solutions are tested across various engines to satisfy all defined criteria.

Constraint extraction automatically identifies all constraints within your problem and ranks them by uniqueness and searchability. This prioritized ranking directs the search orchestration towards the most critical criteria first.

Budgeted, phased validation operates within a 30-search budget across defined phases for extraction, initial search, cross-validation, and revision. Progress is tracked using the [Search X/30] convention until a valid solution is found or the budget is exhausted.

When the 30-search budget is exhausted, the phased validation process stops. The skill tracks progress with the [Search X/30] convention and terminates once the budget is spent without finding a valid candidate solution.

You need multi-engine search for research questions or optimization tasks where multiple constraints must be simultaneously satisfied. It coordinates engines to gather and validate candidate solutions against all defined criteria.