Loading ToolBench
MCP server for Unreal Engine Blueprint Cracker, providing tools for domain capability discovery, input action management, input mapping context management, and input modifier configuration
Server exposes 19 tools with significant definition gaps. Tool names lack consistent verb-noun structure (e.g., 'describe_toolset', 'call_tool' are meta-tools; domain tools like 'CreateInputAction', 'InspectInputAction' use PascalCase inconsistently). Descriptions are present but generic (avg ~80 chars, below 194-char baseline). Input schemas visible for most tools but lack depth: no enums for constrained values (e.g., ValueType in CreateInputAction should enumerate 'axis2d', 'boolean', etc.), no min/max for numeric params, no format specs. Output schemas completely undocumented, no response structure defined for any tool. Error handling absent: no recovery guidance, no categorization of retryable vs fatal errors. Security concerns: no evidence of permission gates on destructive tools (RemoveInputMapping, DiscardOperationBackup). Composition issues: meta-tools (describe_toolset, call_tool) suggest a generic dispatch pattern rather than direct tool exposure, which complicates LLM reasoning.
Adds an input mapping to a UInputMappingContext
Creates a new UInputAction asset with specified properties
Creates a new UInputMappingContext asset
Discards a recovery operation backup
Retrieves server capabilities
Retrieves domain capabilities, optionally filtered by a specific domain name
Inspects an existing UInputAction asset and returns its properties
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields. Violates pattern:tool and pattern:response-shaper.
Constrained parameters lack enums. ValueType in CreateInputAction should enumerate valid types; Accumulation should list allowed modes. Free-form strings invite hallucinated values.
Destructive tools (RemoveInputMapping, DiscardOperationBackup) lack permission gates and confirmation patterns. No evidence of access control or dry-run support.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | <=2025-11-25 | v2 |
Inspects the modifier chain for an input action or mapping
Inspects an existing UInputMappingContext asset and returns its properties
Lists recovery operations that can be discarded
Lists all supported input modifier kinds in the allow-list
Removes an input mapping from a UInputMappingContext
Sets the modifier chain for a UInputAction
Sets the modifier chain for an input mapping in a UInputMappingContext
Validates an existing UInputAction asset
Validates a modifier chain without applying it
Validates an existing UInputMappingContext asset
Calls a tool within a specified toolset with provided arguments
Describes a toolset by name, returning its schema and available tools
Meta-tools (describe_toolset, call_tool) use generic dispatch instead of direct tool exposure. This forces LLMs to reason about toolset names and nested arguments, increasing error likelihood.
No error handling guidance. Tools lack recovery hints, error categorization, or actionable error messages. LLMs cannot self-correct on failures.