Claude Skill

workflow

Infrastructure skill — multi-step action orchestration: run a sequence of MCP tools in order, passing results between steps. Use when chaining two or more tools into a repeatable pipeline (select → rename → validate → export). Not for single-tool operations or DCC-specific busine

LLM Mart · 0 points · 0 views 0 listing impressions 0 install-command copies

#workflow #automation

Download dcc-mcp-dcc-mcp-core-python_dcc_mcp_core_skills_workflow-8d8c969.zip · 5 KB

Install

skills CLI npx skills add https://github.com/dcc-mcp/dcc-mcp-core/tree/main/python/dcc_mcp_core/skills/workflow
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install dcc-mcp-dcc-mcp-core@llmmart
Git git clone https://github.com/dcc-mcp/dcc-mcp-core.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole dcc-mcp/dcc-mcp-core collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Workflow Orchestration

Multi-step action chaining for DCC pipelines.

Overview

The workflow__run_chain tool lets an agent (or a human) execute a sequence of dcc-mcp-core actions in order. Results from earlier steps flow into later steps via context merging and {key} parameter interpolation.

Example: Select → Rename → Validate → Export

{
  "steps": [
    {
      "label": "List mesh objects",
      "action": "maya_scene__list_objects",
      "params": {"type": "mesh"}
    },
    {
      "label": "Rename with prefix",
      "action": "maya_scene__rename_objects",
      "params": {"prefix": "char_"}
    },
    {
      "label": "Validate naming",
      "action": "maya_pipeline__validate_naming",
      "params": {}
    },
    {
      "label": "Export FBX",
      "action": "maya_scene__export_fbx",
      "params": {"output": "/tmp/export.fbx"},
      "stop_on_failure": true
    }
  ]
}

Error Recovery

If a step fails and stop_on_failure is true, the chain halts immediately and returns the partial results so far, plus the error details. The agent can then use dcc_diagnostics__screenshot or dcc_diagnostics__audit_log to investigate before retrying.

Context Interpolation

Use {key} placeholders in params to inject values from the running context:

{"action": "export_fbx", "params": {"output": "{export_path}"}}

The context starts from the context input, then accumulates each step's context output.

Files (dcc-mcp-core)
  • scripts
    • run_chain.py 10 KB
      Loading…
  • SKILL.md 2.2 KB
    ---
    name: workflow
    description: >-
      Infrastructure skill — multi-step action orchestration: run a sequence of MCP
      tools in order, passing results between steps. Use when chaining two or more
      tools into a repeatable pipeline (select → rename → validate → export). Not
      for single-tool operations or DCC-specific business logic — use a domain skill
      for those.
    license: MIT
    metadata:
      dcc-mcp:
        dcc: python
        recall-context:
          app_type: python
          domain: workflow
          workflow_stage: pipeline
          task_category: automate
        version: "1.0.0"
        layer: infrastructure
        search-hint: "chain, sequence, pipeline, multi-step, orchestration, workflow, batch, run steps, automate"
        tags: "workflow, orchestration, chain, pipeline, automation, infrastructure"
        tools: tools.yaml
    ---
    
    # Workflow Orchestration
    
    Multi-step action chaining for DCC pipelines.
    
    ## Overview
    
    The `workflow__run_chain` tool lets an agent (or a human) execute a sequence
    of dcc-mcp-core actions in order. Results from earlier steps flow into later
    steps via context merging and `{key}` parameter interpolation.
    
    ## Example: Select → Rename → Validate → Export
    
    ```json
    {
      "steps": [
        {
          "label": "List mesh objects",
          "action": "maya_scene__list_objects",
          "params": {"type": "mesh"}
        },
        {
          "label": "Rename with prefix",
          "action": "maya_scene__rename_objects",
          "params": {"prefix": "char_"}
        },
        {
          "label": "Validate naming",
          "action": "maya_pipeline__validate_naming",
          "params": {}
        },
        {
          "label": "Export FBX",
          "action": "maya_scene__export_fbx",
          "params": {"output": "/tmp/export.fbx"},
          "stop_on_failure": true
        }
      ]
    }
    ```
    
    ## Error Recovery
    
    If a step fails and `stop_on_failure` is `true`, the chain halts immediately
    and returns the partial results so far, plus the error details. The agent can
    then use `dcc_diagnostics__screenshot` or `dcc_diagnostics__audit_log` to
    investigate before retrying.
    
    ## Context Interpolation
    
    Use `{key}` placeholders in `params` to inject values from the running context:
    
    ```json
    {"action": "export_fbx", "params": {"output": "{export_path}"}}
    ```
    
    The context starts from the `context` input, then accumulates each step's
    `context` output.
    
  • tools.yaml 1.4 KB
    Loading…

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.