Replaces operationId-based tool naming with a deterministic path-based naming engine.

How it works

Each path is parsed into typed components (RESOURCE or PARAM), then processed to produce a camelCase tool name:

/repos/{owner}/{repo}/issues/{id}
   |     |       |      |      |
   |    [params - skipped]     |
   |                           |
container                   resource
(skipped)                 singularized
                        -> "getRepoIssue"

A container is a resource that scopes other resources through path parameters.. it's dropped since it does not describe the operation. The remaining resource chain is singularized or pluralized based on position and HTTP method, then abbreviated where applicable.

This generalizes across API conventions:

Changes

• Adds inflector.py.. English pluralization/singularization for API resource names
• Adds tool_naming.py.. ToolNamingEngine generates consistent camelCase tool names from REST paths
• Adds abbrs.json (mirrored from ocp-registry).. abbreviation dictionary used by the naming engine
• Updates openapi_parser.py to use ToolNamingEngine instead of normalizing operationId