本文へ移動
cccskills
無料GitHub で公開

bdb-touchdesigner-mcp

Utilizes the TouchDesigner MCP servers (tdmcp and touchdesigner-mcp) to create TOP/CHOP chains, script operators, inspect parameters, and debug node networks.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md6.5 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

TouchDesigner MCP — Integration and AI Agent Guide

This skill file instructs AI agents on how to construct real-time visual systems inside TouchDesigner using the tdmcp (MindDesigner) and touchdesigner-mcp (Backup) integrations. It details node building, operator classes, preview generation, and troubleshooting steps.

1. Overview and Pipeline Value

The TouchDesigner MCP (MindDesigner) server enables AI models to programmatically construct, connect, parameterize, and debug node networks in TouchDesigner. In the BDB OS architecture, this automates audio-reactive particle simulations, feedback systems, projection mapping, GLSL shader chains, and data visualizations.

Architecture

  • In-Process WebServer / Socket Bridge: A custom component (e.g. tdmcp_bridge_package.tox or mcp_webserver_base.tox) is placed inside TouchDesigner. It listens on a local port and translates incoming JSON commands to Python operations (like parent().create(noiseTOP)).
  • Dual Server Layout:
    • tdmcp (MindDesigner): Runs on port 9980. Focuses on creative visual engineering with 332 high-level generation tools.
    • touchdesigner-mcp (Backup/Direct): Runs on port 9981. Focuses on atomic node CRUD and low-level scripting fallback.

Which server

ServerUse forFallback when
bdb_td_minddesigner (tdmcp, port 9980)Primary: high-level creative generation and structured inspection/edit tools.Undocumented.
bdb_td_backup (touchdesigner-mcp, port 9981)Atomic node CRUD and low-level scripting.bdb_td_minddesigner is unavailable (pair labelled primary/fallback; no other trigger documented).

2. System Instructions

Workflow Priorities

  1. Bridge Verification first: Always call get_td_info to confirm the TouchDesigner bridge is reachable and running compatible versions.
  2. Consult Operator Knowledge: Consult the operator library before placing nodes. Never guess operator types or namespaces (e.g. use noiseTOP for a Noise TOP, constantCHOP for a Constant CHOP).
  3. Structured Tools first: Prefer structured inspection and modification tools (find_td_nodes, get_td_node_parameters, update_td_node_parameters, get_td_node_errors) over raw Python execution. Use execute_python_script only when no structured tool fits the request.
  4. Visual Feedback: After creating or modifying a network, call get_preview or get_inline_preview so the user can verify the rendering in real-time.

3. Available Tools and API Parameters

High-Level Creative Tools (tdmcp)

  • create_audio_reactive(layer_name: string): Sets up a complete audio spectrum analyzer routing to video modifiers.
  • create_feedback_network(parent: string, input_op: string): Creates a feedback loop using feedback, blur, level, and composite TOPs.
  • create_particle_system(parent: string): Generates a GPU particle flow using SOPs, CHOPs, and TOPs.
  • animate_parameter(op_path: string, par_name: string, chop_expr: string): Binds a node parameter to a CHOP expression.

Atomic Node Tools (touchdesigner-mcp & fallback)

  • create_td_node(type: string, path: string, name: string): Places a single node in a network container.
  • delete_td_node(path: string): Deletes an operator.
  • connect_nodes(from_op: string, to_op: string, from_connector?: int, to_connector?: int): Connects outputs to inputs.
  • get_td_nodes(path: string): Lists all operators in a specific network.
  • get_td_node_errors(path: string): Checks for active operator errors (red flags) recursively.
  • update_td_node_parameters(path: string, parameters: object): Modifies parameter settings.

4. Code Recipes and Prompt Cookbook

Recipe 1: Procedural Feedback Loop (MindDesigner)

Generate a feedback warp tunnel TOP chain:

// Step 1: Create noise source
// Tool: create_td_node(type="noiseTOP", path="/project1", name="noise_src")

// Step 2: Set noise parameters
// Tool: update_td_node_parameters(path="/project1/noise_src", parameters={"resolutionw": 1080, "resolutionh": 1080, "period": 2.5})

// Step 3: Set up feedback loop
// Tool: create_feedback_network(parent="/project1", input_op="/project1/noise_src")

Recipe 2: Run Custom Python Script for Network Connection

Connect a series of CHOPs programmatically when structured connection tools are not flexible enough:

# Call tool: execute_python_script(script="...")
parent_comp = op('/project1')
lfo = parent_comp.create(lfoCHOP, 'oscillator')
math_op = parent_comp.create(mathCHOP, 'modifier')
null_op = parent_comp.create(nullCHOP, 'out_signal')

# Connect ports
math_op.inputConnectors[0].connect(lfo)
null_op.inputConnectors[0].connect(math_op)

# Set Math CHOP range multiplier
math_op.par.mult = 100.0

5. Troubleshooting and Connection Details

Configuration and Ports

  • MindDesigner Port: 9980 (TDMCP_PORT environment variable)
  • Backup Controller Port: 9981 (WebServer port)
  • Verify Endpoint Availability:
    # Check if MindDesigner is reachable
    curl http://localhost:9980/api/info
    

Bridge Installation inside TouchDesigner

  1. Download tdmcp_bridge_package.tox or mcp_webserver_base.tox.
  2. Drag the .tox component directly into your active network pane (e.g. /project1).
  3. Press Install / verify the active port is running.

Alternative Textport Bootstrap Installation: Open TouchDesigner Textport (Dialogs -> Textport and DATs), paste and run:

import urllib.request; exec(urllib.request.urlopen("https://github.com/Pantani/tdmcp/raw/v0.12.0/td/bootstrap.py").read().decode())

Common Errors and Fixes

  1. TouchDesignerClient ECONNREFUSED
    • Cause: The bridge component is either not installed or disabled in TouchDesigner.
    • Fix: Ensure you dragged the .tox component into the project and that its internal server is toggled active.
  2. Operator compilation error or red flag
    • Cause: An invalid parameter type or code was set.
    • Fix: Run get_td_node_errors on the path of the flagged node to read the compiled compiler/interpreter traceback.
  3. Version Mismatch Warning
    • Fix: Download the latest bridge zip file from releases, delete the old touchdesigner-mcp-td folder in your project, restart TouchDesigner, and drop in the new .tox.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Use when manage multiple local CLI agents via tmux sessions (start/stop/monitor/assign) with cron-friendly scheduling.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Use when a hybrid memory system that provides persistent, searchable knowledge management for AI agents (Architecture, Patterns, Decisions).

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Reference for how BDB structures autonomous software engineering work — the seven-node dispatcher graph (Architect, TechLead, UI/UX, Engineering, Media/EventTech, Reviewer, Shipping) that /startcycle-graph actually runs. Use when you need the high-level lifecycle framing without inventing your own process.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Tools are how AI agents interact with the world. A well-designed tool is the difference between an agent that works and one that hallucinates, fails silently, or costs 10x more tokens than necessary. This skill covers tool design from schema to error handling.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Harness patterns for coding agents — memory, permissions, context engineering, delegation, skills, hooks, bootstrap.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

agenttrail: live map of a multi-agent build in the browser: which plan component is being worked on, by which agent or harness, what is done and what is stuck. Use when a multi-agent pipeline starts (/startcycle, /startcycle-graph, /teamwork-preview) or after a plan-canvas approve, when the user asks to see what the agents are doing ("live map", "build board", "who is running now"), or by yourself whenever a multi-agent run is under way — no user prompt needed.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

hybridlabor-api のスキルをすべて見る

このスキルの問題を報告する