Six tools with basic schemas and descriptions, but significant gaps in parameter documentation, error handling, and composition. Tool names lack clear action verbs (clip_control, layer_control, effect_control are vague). Descriptions are present but generic (avg ~60 chars, below 194-char baseline). Most parameters lack descriptions entirely. No output schemas documented. No error recovery guidance. No idempotency or composition hints. The 'help' tool is a red flag, agents should not need to call a help tool; tool descriptions should be self-documenting.

• Rename tools with clear action verbs: 'load_clip', 'trigger_clip', 'set_clip_position', 'set_clip_opacity', 'set_layer_opacity', 'set_layer_blend_mode', 'set_layer_transition', 'set_effect_parameter', 'bypass_effect', 'set_bpm', 'get_performance_status'. Each name should start with a verb and convey the exact action.
• Add descriptions to every parameter. For 'value', specify: 'Numeric value for the property. For opacity/position: 0.0 - 1.0 (0=transparent/left, 1=opaque/right). For blend_mode: 0 - N (see Resolume blend mode enum). For transition: seconds (0 - 60).'
• Document output schemas for each tool. E.g., 'load_clip returns {clip_id: string, layer: int, name: string, duration_ms: int}' so agents can chain calls and extract IDs.
• Add error handling with recovery hints. E.g., 'If layer index is out of range, returns error: "Layer 99 not found. Available layers: 1 - 8. Try a valid layer number."'
• Remove the 'help' tool. Expand tool descriptions to 100 - 200 chars each, covering WHAT, WHEN, and WHAT IT RETURNS. Include dependency hints like 'Call status() first to see available layers.'