PyMOL-MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PYMOL_MCP_HISTORY | No | Directory for session history files, or 'off' to disable. Default is ~/.pymol-mcp/. Set to a custom path or 'off'. | |
| PYMOL_MCP_LOG_LEVEL | No | Log level for the MCP server. Set to 'INFO' or 'DEBUG' to enable logging. Default is no logging. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| parse_and_executeA | Executes a single PyMOL command given in literal PyMOL syntax. NOT a natural-language interface.
Translating requests: "Load PDB 1UBQ and show it as cartoon" -> parse_and_execute("fetch 1ubq") -> parse_and_execute("as cartoon, 1ubq") "Colour chain A red" -> "color red, chain A" "Show sticks for residues 1-50" -> "show sticks, resi 1-50" "Open /data/model.pdb" -> "load /data/model.pdb" "Select the binding site" -> "select site, byres (polymer within 5 of ligand)" Common mistakes:
Selections use full PyMOL algebra ( Returns PyMOL's output, or a message describing the parse/execution failure. |
| list_instancesA | Lists the running PyMOL instances and what each has loaded. Each PyMOL claims its own port, so several can run at once. Pass a port as
The loaded object names are what distinguish one window from another; a port number on its own identifies nothing to a human. |
| list_commandsA | Lists the PyMOL commands Without Examples: filter="color" for the colouring commands, filter="cartoon" for cartoon-related ones, filter="fetch" for the exact fetch signature. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 3 tools
Each tool serves a distinct purpose: list_commands provides command syntax, list_instances shows running instances, and parse_and_execute executes commands. There is no overlap.
All tool names use lowercase snake_case with a verb_noun pattern (list_commands, list_instances, parse_and_execute). The naming is consistent and predictable.
With 3 tools, the set is well-scoped for the server's purpose—providing help, instance information, and command execution. No unnecessary tools.
The tool surface covers the core workflow of querying syntax, checking instances, and executing commands. A minor gap is the lack of a direct way to get the current state or result of previous commands, but agents can work around this.