Tool definitions are present with schemas and descriptions, but quality is inconsistent. Most tools have basic descriptions (50-150 chars) and parameter schemas with types, but lack depth in error guidance, output documentation, and LLM-optimized phrasing. No tool annotations (readOnlyHint/destructiveHint) despite clear risk classifications. Parameter descriptions are minimal (10-30 chars). Output schemas are not documented. Error handling is absent from definitions. Naming follows verb_noun convention (asset.import, capture.scene) but some names are ambiguous (generator.quote vs generator.submit). No evidence of idempotency guarantees or recovery guidance.

• Add tool annotations to all definitions: mark asset.import, asset.create, checkpoint.create, console.clear, generator.submit as destructiveHint=true; mark capture.scene, capture.game, console.get_logs, checkpoint.list as readOnlyHint=true; mark checkpoint.restore as idempotentHint=true if it is.
• Document output schemas for every tool. Example: 'Returns {asset_id: string, path: string, created_at: ISO8601}'. This enables LLMs to chain tools and extract required fields.
• Expand parameter descriptions to 50-100 chars with constraints. For asset.create 'type', change to: 'Asset type to create. Must be one of: material, script, shader, animator, scriptableobject, prefab, scene.'
• Add recovery guidance to error-prone tools. For asset.import: 'If import fails, verify the source file exists and is in a supported format (FBX, PNG, etc.). Call asset.list to see imported assets.'
• Clarify generator workflow: rename 'generator.quote' to 'generator.estimate' or add description: 'Previews generation cost and time without executing. Call generator.submit to start the job.'