CatGO Workflow Skill

SkillDev tools

Create and manage computational chemistry workflows with CatGo. Supports VASP, CP2K, ORCA, MLP, LAMMPS. Build OER/HER/CO2RR workflows, geometry optimization, frequency analysis, Gibbs energy calculations.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the CatGO Workflow Skill skill

What this skill tells your AI

The instructions your AI receives, as published by hello-qm/catgo-lrg in server/catgo/workflow/SKILL.md and read by ahel’s review.

All MCP workflow operations use the single catgo_workflow tool with an action parameter.

Current Reality

Do not assume old workflow docs are still correct.

Important current behavior:

  • create creates a workflow that already contains one structure_input node.
  • run starts execution immediately. It does not open a confirmation dialog.
  • connect defaults both handles to structure if omitted.
  • node_types is only a discovery list. It does not give full node schema.

Quick Reference

ActionPurposeRequired Params
listList all workflows
templatesList available templates
node_typesList node types by categorycategory?
createCreate workflowname, template_id?
getRead workflow graphworkflow_id
add_nodeAdd nodeworkflow_id, node_type, params?
remove_nodeRemove nodeworkflow_id, node_id
connectConnect nodesworkflow_id, from_id, to_id, from_handle?, to_handle?
set_paramsUpdate paramsworkflow_id, node_id, params
validateValidate graphworkflow_id
runStart execution immediatelyworkflow_id, run_config?
pausePause workflowworkflow_id
resumeResume workflowworkflow_id, run_config?
statusGet run statusworkflow_id
step_errorGet failure detailworkflow_id, step_id

Safe Build Sequence

1. Discover

  • Call node_types first for coarse discovery.
  • Call templates if the user wants a preset.

2. Create

{"action": "create", "name": "TiO2 slab workflow"}

Immediately after create:

  • Call get
  • Check the graph
  • Reuse the auto-created structure_input unless the workflow truly needs multiple independent inputs

3. Add Nodes

{"action": "add_node", "workflow_id": "...", "node_type": "geo_opt"}
{"action": "add_node", "workflow_id": "...", "node_type": "single_point"}

4. Set Parameters

{"action": "set_params", "workflow_id": "...", "node_id": "...", "params": {"software": "vasp", "ENCUT": 520}}

5. Connect Carefully

Only omit handles when both sides are obviously single-structure nodes.

Safer form:

{
  "action": "connect",
  "workflow_id": "...",
  "from_id": "n1",
  "to_id": "n2",
  "from_handle": "structure",
  "to_handle": "structure"
}

Do Not Do These

  • Do not assume create returns an empty graph.
  • Do not blindly add another structure_input after create.
  • Do not assume connect can guess the right handle for multi-port nodes.
  • Do not treat validate warnings as harmless.
  • Do not call run before the user has provided a valid run configuration when HPC settings matter.

Minimal Reliable Pattern

  1. node_types
  2. create
  3. get
  4. add_node
  5. set_params
  6. connect with explicit handles when needed
  7. validate
  8. get
  9. run

Signals

GitHub stars
196
Forks
23
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
workflow-hello-qm
Source
github.com/hello-qm/catgo-lrg