tdmcp

Create slit-scan

create_slit_scan

Creates a slit-scan visual system in TouchDesigner, sampling past frames per row/column for a time-stretch effect. Includes cache ring buffer, shader, and optional synthetic source for headless use.

Instructions

Build a slit-scan visual system: each row (or column) of the output samples a different past frame from a Cache TOP ring buffer, producing the classic 'time-as-space' stretched-time look (Floris Kaayk / Adam Magyar style). Creates a new baseCOMP under parent_path holding a source TOP, a Cache TOP ring buffer, a slit GLSL shader, and a Null output. When no source_top_path is given, a synthetic Noise TOP is used so the tool works headless / on CI without camera permission. Exposes a live 'Depth' knob. Note: GLSL compile is UNVERIFIED offline; cacheTOP 2D-array binding must be validated live in TouchDesigner. Memory cost at 1080p RGBA16 is ~depth × 32 MB; depth 600 ≈ 5 GB VRAM. Output freezes when the timeline is paused (cacheTOP stops recording — expected behaviour). Returns a summary plus a JSON block with the container path, created node paths, output path, exposed controls, warnings, and an inline preview image.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "create_slit_scan"
  }
}

Input Schema

Name	Required	Description	Default
`axis`	No	Screen axis that carries time. 'y' = each row is a different past frame; 'x' = each column (default 'y').	y
`name`	No	Container name for the slit-scan system (default 'slit_scan').	slit_scan
`direction`	No	Which end of the axis is 'now'. '+y' = bottom row is the latest frame, top is oldest; '-y' reverses it. Must be compatible with axis (default '+y').	+y
`cache_depth`	No	Number of frames stored in the Cache TOP ring buffer (1–600, default 60). Memory cost: ~depth × W × H × 16 B at RGBA16. At 1080p, 600 frames ≈ 5 GB VRAM.
`parent_path`	No	Parent network where the slit-scan container is created (default '/project1').	/project1
`expose_controls`	No	When true (default), expose a live 'Depth' knob on the container bound to cache.cachesize.
`source_top_path`	No	Optional path to an existing TOP to scan (e.g. '/project1/videodevicein1'). When omitted a synthetic noiseTOP seed is created inside the container so the tool runs headless / on CI without camera permission.

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.13.1Aug 2, 2026

TDQS

A4.5/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Goes well beyond annotations by disclosing critical limitations: GLSL compile is unverified offline, cacheTOP binding must be validated live, memory cost scaling (depth × 32 MB at 1080p), and output freezes when timeline pauses. This is exactly the kind of behavioral nuance an agent needs and adds substantial value over the annotations, which only state it is not read-only and not destructive.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is dense yet efficiently structured. Every sentence delivers specific, non-redundant information: mechanism, components, headless behavior, exposed knob, verification warnings, memory costs, pause behavior, and return value. It is front-loaded with the core purpose and maintains clarity throughout.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a tool with 7 parameters and no output schema, the description is exceptionally complete. It explains the internal architecture, return value shape, runtime pitfalls, and resource requirements. No significant gaps remain for an agent to safely invoke and interpret results.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so parameters are already well-documented. The description adds a bit of extra context (e.g., 'Depth' knob binding to cache.cachesize, memory formula for cache_depth) but mostly repeats or paraphrases schema descriptions. It does not significantly deepen understanding beyond the schema, hence baseline 3.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a clear, specific verb ('Build a slit-scan visual system') and explains the mechanics (row/column samples past frames from a Cache TOP ring buffer). It distinguishes itself from sibling creation tools by naming the exact technique and visual style (Floris Kaayk / Adam Magyar), making its purpose unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage context is clearly described: it is the tool for creating slit-scan looks, and it explicitly notes it works headless/CI when no source_top_path is given. However, it does not name alternative tools for similar effects (e.g., create_time_echo), so it lacks explicit when-not-to-use guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server