Loading ToolBench
MCP Server for computing fair surplus/cost allocation in cooperative games using Shapley value calculation. Implements the uniquely fair allocation satisfying Efficiency, Symmetry, Dummy Player, and Additivity axioms across all player coalitions.
Single tool with adequate schema but weak descriptions and no error handling. Tool name lacks action verb (should be 'compute_shapley_allocation'). Description is 194 chars (within baseline 34-392 range) but lacks WHEN to use it and prerequisites. Input schema present with types and defaults, but parameter descriptions are minimal (10-30 chars, well below 72-char baseline). No output schema documented. No error guidance, validation rules, or recovery paths. No security considerations despite handling numeric allocations.
Compute exact Shapley value fair allocation for players in a cooperative game. Takes a list of players, their standalone values, and a grand coalition bonus, then calculates the uniquely fair surplus allocation satisfying Efficiency, Symmetry, Dummy Player, and Additivity axioms.
Tool name lacks action verb. 'calculate_shapley_allocation' is acceptable but 'compute_shapley_allocation' (already in description) would be clearer. Current name is passive-leaning.
Parameter descriptions are too brief (10-30 chars vs 72-char baseline). 'List of player identifiers (strings)' lacks format constraints, valid ranges, or examples. 'Dictionary mapping each player to their standalone value (float)' does not explain what 'standalone value' means in game theory context.
No output schema documented. Tool returns {'shapley_values': shares} but LLM cannot infer field types, ranges, or structure. Should document: 'Returns dict mapping player names to float allocations summing to grand_coalition_value.'
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 44 | 2026-07-28+ | v2 |
No error handling or validation. Tool does not validate: player list non-empty, standalone_values keys match players, numeric values are finite, grand_coalition_bonus >= 0. Invalid input silently produces wrong results instead of actionable errors.
Tool description lacks WHEN to use it and prerequisites. Does not explain: what Shapley values are, when fair allocation is needed, what axioms are satisfied, or what assumptions the characteristic function must meet.