feat(ai): add Unity gameplay authoring and MCP workflows

[marcosotomac]

Summary

This PR introduces a complete Unity AI authoring workflow focused on reusable gameplay templates, MCP capabilities, content importing, runtime script authoring, scene diagnostics, physics/input workflows, and asset telemetry.

The goal is to make the Unity AI system more reliable, extensible, and useful for creating, importing, auditing, and validating gameplay content directly through structured AI-assisted workflows.

Main Changes

Unity AI gameplay authoring

• Added reusable gameplay templates for faster scene and gameplay generation.
• Expanded Unity content import and audit reporting.
• Added guarded runtime script authoring to make generated scripts safer and easier to validate.
• Exposed render and input environment capabilities.
• Added support for stable package asset metadata.

MCP and bridge improvements

• Exposed content and gameplay capabilities through MCP.
• Published complete local MCP tool schemas.
• Improved bridge resilience during Unity domain reloads.
• Recovered missing desktop session state.
• Added missing desktop session handling for more stable workflows.

Assets and catalog workflows

• Imported verified catalog content.
• Added end-to-end catalog import verification.
• Expanded starter catalog aliases.
• Documented resilient bridge and asset catalog workflows.
• Added generic asset telemetry workflows.
• Added normalization settings telemetry for control/configuration flows.

Scene, physics, and input workflows

• Added bounded hierarchy filters for scene inspection.
• Bound serialized object references for scene workflows.
• Added bounded scene diagnostics for physics.
• Preserved Unity 2022 velocity API compatibility.
• Simulated Input System events for testing.
• Documented scene physics and input workflows.

Console, observer, and quality improvements

• Isolated blocking compilation errors in the console.
• Tolerated omitted radius filters in observer flows.
• Added UI quality screens for auditing and validation.
• Improved test coverage for advanced content workflows.
• Added end-to-end tests for scene physics and console upgrades.

Testing

• Added tests for advanced content workflows.
• Added tests for catalog import end-to-end verification.
• Added tests for scene physics and console upgrades.
• Added Input System event simulation coverage.

Value

This PR strengthens the Unity AI project by making gameplay authoring more structured, testable, and reusable. It also improves MCP integration, scene analysis, content import reliability, runtime script safety, and asset telemetry workflows, which are key pieces for building a more complete AI-assisted Unity development environment.

Notable Commits Covered

• feat(unity): expand content import and audit reporting
• feat(unity): add guarded runtime script authoring
• feat(unity): compose reusable gameplay templates
• feat(mcp): expose content and gameplay capabilities
• fix(mcp): publish complete local tool schemas
• fix(bridge): survive Unity domain reloads
• fix(bridge): recover missing desktop session state
• feat(scene): add bounded hierarchy filters
• feat(scene): bind serialized object references
• feat(physics): add bounded scene diagnostics
• fix(physics): preserve Unity 2022 velocity API
• feat(tests): simulate Input System events
• fix(console): isolate blocking compilation errors
• feat(ui): compose and audit quality screens
• feat(assets): expand starter catalog aliases
• feat(control): add generic normalization settings telemetry
• docs(control): document generic asset telemetry workflows