Build, run, and optimize composable cheminformatics and computational chemistry pipelines using an LLM agent.
Also known as b-shields/cmxflow
- Stars
- 2
- Tools
- —
- Downloads, 30 days
- 94
- Latest version
- 0.4.0
- Last commit
- Aug 31, 2026
- License
- MIT
What is cmxflow MCP?
cmxflow is an MCP server. Build, run, and optimize composable cheminformatics and computational chemistry pipelines using an LLM agent.
Summary generated from the project's README
About cmxflow MCP server
cmxflow provides a framework for constructing and managing cheminformatics and computational chemistry pipelines through composable blocks. It is designed for researchers and developers who need to automate molecular workflows such as ligand preparation, docking, and virtual screening. Users can build these pipelines manually using Python or interactively through an LLM agent using the provided MCP server.
The server includes five specific tools: build_workflow, run_workflow, optimize_workflow, manage_workflows, and view_structures. It supports Bayesian optimization via Optuna to tune parameters like fingerprint types or similarity metrics. For 3D visualization of structures, the view_structures tool requires PyMOL to be installed. The system handles compute-heavy tasks like conformer generation and docking through parallel execution and offers scaffold-indexed docking modes to increase throughput on congeneric series.
What you can do with it
- 1Automate ligand preparation workflowsCreate a pipeline that standardizes, ionizes, and generates 3D conformers for a library of molecules.
- 2Perform molecular docking tasksDock a set of molecules against a specific protein receptor using a reference ligand.
- 3Optimize virtual screening parametersUse Bayesian optimization to find the best fingerprint and similarity settings to maximize enrichment scores.
- 4Analyze molecular properties conversationallyAsk an AI agent to filter a chemical library based on specific medicinal chemistry rules.
Written from the project's README. Every use case is checked against a quote from it.
How to install cmxflow MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall cmxflow MCP in Claude Code
$ claude mcp add cmxflow -- uvx cmxflowInstall cmxflow MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"cmxflow": {
"command": "uvx",
"args": [
"cmxflow"
]
}
}
}Install cmxflow MCP in VS Code
.vscode/mcp.json
{
"servers": {
"cmxflow": {
"type": "stdio",
"command": "uvx",
"args": [
"cmxflow"
]
}
}
}Install cmxflow MCP in VS Code CLI
$ code --add-mcp '{"name":"cmxflow","type":"stdio","command":"uvx","args":["cmxflow"]}'Install cmxflow MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"cmxflow": {
"command": "uvx",
"args": [
"cmxflow"
]
}
}
}Install cmxflow MCP in Codex
$ codex mcp add cmxflow -- uvx cmxflowInstall cmxflow MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.cmxflow] command = "uvx" args = ["cmxflow"]
Install cmxflow MCP in Gemini CLI
$ gemini mcp add -s user cmxflow uvx -- cmxflowInstall cmxflow MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"cmxflow": {
"command": "uvx",
"args": [
"cmxflow"
]
}
}
}Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachableb-shields/cmxflow is public and reachablegithub.com/b-shields/cmxflow ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Sigstore attestation for cmxflowpypi.org/project/cmxflow/ ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-08-31github.com/b-shields/cmxflow ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from cmxflowPass
- No critical advisory in the package or its dependencies91 resolved dependencies checked; no critical advisoryosv.dev ↗Pass
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubcmxflow 🧪
Build cheminformatics and computational chemistry pipelines with composable blocks. Tune end-to-end with Bayesian Optimization. Or ask an LLM agent to do it.
Quick examples
Prepare ligands for docking
from cmxflow import Workflow
from cmxflow.sources import MoleculeSourceBlock
from cmxflow.operators import (
MoleculeStandardizeBlock,
IonizeMoleculeBlock,
EnumerateStereoBlock,
ConformerGenerationBlock,
)
from cmxflow.sinks import MoleculeSinkBlock
# Standardize → ionize (pH 6.4–8.4) → enumerate stereo → generate 3D conformers
workflow = Workflow()
workflow.add(
MoleculeSourceBlock(),
MoleculeStandardizeBlock(),
IonizeMoleculeBlock(),
EnumerateStereoBlock(),
ConformerGenerationBlock(),
MoleculeSinkBlock(),
)
workflow("library.smi", "prepared.sdf")
Dock a congeneric series
Pure-Python docking. Free docking is the default (index_poses=False); scaffold-indexed mode caches poses by Bemis–Murcko scaffold for ~3× faster throughput on congeneric series with consistent pose alignment.
from cmxflow import Workflow
from cmxflow.sources import MoleculeSourceBlock
from cmxflow.operators import ConformerGenerationBlock, MoleculeDockBlock
from cmxflow.sinks import MoleculeSinkBlock
from cmxflow.utils.parallel import make_parallel
workflow = Workflow()
workflow.add(
MoleculeSourceBlock(),
ConformerGenerationBlock(),
make_parallel(
MoleculeDockBlock(
receptor="receptor.pdb",
site_reference="crystal_ligand.sdf",
index_poses=True, # omit for free docking
)
),
MoleculeSinkBlock(),
)
workflow("library.smi", "docked.sdf")
Tune a ligand-based virtual screen
from cmxflow import Workflow
from cmxflow.sources import MoleculeSourceBlock
from cmxflow.operators import MoleculeSimilarityBlock
from cmxflow.scores import EnrichmentScoreBlock
from cmxflow.opt import Optimizer
# Rank a library by 2D similarity to a known active, then tune the
# fingerprint end-to-end to maximize enrichment AUC.
workflow = Workflow()
workflow.add(
MoleculeSourceBlock(),
MoleculeSimilarityBlock(queries="crystal_ligand.sdf"),
EnrichmentScoreBlock(target="active"),
)
opt = Optimizer(workflow, "benchmark.csv")
opt.optimize(n_trials=30, direction="maximize")
print(f"Best enrichment AUC: {opt.best_score:.3f}")
print(opt.best_params)
# Best enrichment AUC: 0.836
# {'fingerprint_type': 'morgan', 'similarity_metric': 'sokal', 'radius': 2, 'nbits': 2545}
The four fingerprint parameters above are searched automatically — every block exposes its mutable parameters to the optimizer.
Or build it conversationally via an LLM agent
claude mcp add cmxflow -- cmxflow-mcp
"How many of the molecules in library.csv pass Lipinski's rules?"
"I need to build a ligand-based virtual screening workflow. I'm not sure if 2D or 3D is better. Can you optimize two workflows?"
"Dock the molecules in hits.csv against receptor.pdb with crystal_ligand.sdf as a reference."
The agent can build, run, and optimize workflows. See Using with Claude for full transcripts.
What's in the box
- 15+ blocks for sourcing, transforming, filtering, clustering, scoring, and docking molecules
- Bayesian optimization of pipeline parameters via Optuna
- Parallel execution for compute-heavy blocks (conformer generation, docking)
- Workflow serialization for save / load / reuse
- An MCP server with five tools:
build_workflow,run_workflow,optimize_workflow,manage_workflows,view_structures
Install
pip install cmxflow
MCP server
claude mcp add cmxflow -- cmxflow-mcp
Optional: PyMOL
Required only for the view_structures MCP tool (3D visualization):
conda install -c conda-forge pymol-open-source
Documentation
- Docs site
- Block catalog
- Using with Claude — agent transcripts
examples/basic_usage.ipynb— full tutorialexamples/docking/docking.ipynb— docking walkthrough (ILS, scaffold-indexed, and template modes)
Project
MIT licensed. See CONTRIBUTING.md and RELEASING.md.
cmxflow MCP questions
How do I add cmxflow MCP to Claude Code?
Run claude mcp add cmxflow -- uvx cmxflow in your terminal. Add --scope user to make it available in every project.
How do I add cmxflow MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
What tools are available in the MCP server?
The MCP server provides five specific tools: build_workflow, run_workflow, optimize_workflow, manage_workflows, and view_structures.
How does the docking process handle speed for similar molecules?
It offers a scaffold-indexed mode that caches poses by Bemis–Murcko scaffold, which can be approximately 3x faster for congeneric series.