molclaw-chroma-toolkit

name	molclaw-chroma-toolkit
description	Chroma toolkit skill covering chroma_monomer for single-chain generation, chroma_complex for multi-chain assembly generation, and chroma_symmetry for symmetry-constrained protein design.
license	MIT license
metadata	{"skill-author":"PJLab","tool-summary":"chroma_monomer: Generates a de novo single-chain protein candidate.\nchroma_complex: Generates a de novo multi-chain protein complex candidate.\nchroma_symmetry: Generates a symmetry-constrained oligomeric protein candidate.\n"}

Chroma Protein Generation Toolkit

Note:

• Local files are not directly accessible by the server. Please upload them to the server using drugsda-file-transfer before execution.
• For PDB file inputs, it is recommended to preprocess them using drugsda-fix_pdb before execution.

Usage

2. Monomer Protein Generation

The description of tool chroma_monomer.

Generates a de novo single-chain protein candidate for exploratory protein design workflows.
Args:
    length (int): Number of residues in the monomer chain. Default: 100.
    steps (int): Diffusion sampling step count. Default: 500.
    device (str): Chroma execution device string. Default: 'cuda:0'.
    fmt (str): Output structure format, one of {'cif', 'pdb'}. Default: 'cif'.
    api_key (str|None): Optional Chroma API key for model access.
    dry_run (bool): If True, validate parameters and prepare output paths without model sampling. Default: False.
Return:
    status (str): 'success' or 'error'.
    msg (str): Human-readable execution summary.
    tool_name (str): Tool identifier 'chroma_monomer'.
    output_dir (str): Run-specific output directory under tool_result/chroma_toolkit_result.
    output_file (str): Target output structure file path.
    length (int): Effective residue length used in generation.
    steps (int): Effective sampling steps used in generation.
    device (str): Effective execution device.
    format (str): Effective output format.

How to use tool chroma_monomer :

response = await client.session.call_tool(
    "chroma_monomer",
    arguments={
        "length": 150,
        "steps": 500,
        "device": "cuda:0",
        "fmt": "cif",
        "dry_run": False
    }
)
result = client.parse_result(response)
output_file = result["output_file"]

Example parameter sets

# 1) Main mode
{
    "length": 150,
    "steps": 500,
    "device": "cuda:0",
    "fmt": "cif",
    "dry_run": False
}

# 2) Variant mode
{
    "length": 100,
    "steps": 50,
    "fmt": "cif",
    "dry_run": True
}

3. Complex Protein Generation

The description of tool chroma_complex.

Generates a multi-chain protein complex candidate for interface and assembly design studies.
Args:
    chains (str): Comma-separated chain lengths, for example '100,120'. Default: '100,100'.
    steps (int): Diffusion sampling step count. Default: 500.
    device (str): Chroma execution device string. Default: 'cuda:0'.
    fmt (str): Output structure format, one of {'cif', 'pdb'}. Default: 'cif'.
    api_key (str|None): Optional Chroma API key for model access.
    dry_run (bool): If True, validate parameters and prepare output paths without model sampling. Default: False.
Return:
    status (str): 'success' or 'error'.
    msg (str): Human-readable execution summary.
    tool_name (str): Tool identifier 'chroma_complex'.
    output_dir (str): Run-specific output directory under tool_result/chroma_toolkit_result.
    output_file (str): Target output structure file path.
    chains (str): Effective chain-length specification used in generation.
    steps (int): Effective sampling steps used in generation.
    device (str): Effective execution device.
    format (str): Effective output format.

How to use tool chroma_complex :

response = await client.session.call_tool(
    "chroma_complex",
    arguments={
        "chains": "100,120",
        "steps": 500,
        "device": "cuda:0",
        "fmt": "cif",
        "dry_run": False
    }
)
result = client.parse_result(response)
output_file = result["output_file"]

Example parameter sets

# 1) Main mode
{
    "chains": "100,120",
    "steps": 500,
    "device": "cuda:0",
    "fmt": "cif",
    "dry_run": False
}

# 2) Variant mode
{
    "chains": "80,80,80",
    "steps": 50,
    "fmt": "pdb",
    "dry_run": True
}

4. Symmetry-Constrained Protein Generation

The description of tool chroma_symmetry.

Generates a symmetry-constrained protein design candidate for oligomeric architecture exploration.
Args:
    group (str): Symmetry group label such as 'C_3' or 'D_2'.
    length (int): Residue count per protomer chain. Default: 100.
    steps (int): Diffusion sampling step count. Default: 500.
    num_chain_neighbors (int): Neighbor-chain count for symmetry conditioner. Default: 2.
    langevin_factor (float): Langevin factor for conditioned sampling. Default: 8.0.
    inverse_temperature (float): Inverse temperature for conditioned sampling. Default: 8.0.
    device (str): Chroma execution device string. Default: 'cuda:0'.
    fmt (str): Output structure format, one of {'cif', 'pdb'}. Default: 'cif'.
    api_key (str|None): Optional Chroma API key for model access.
    dry_run (bool): If True, validate parameters and prepare output paths without model sampling. Default: False.
Return:
    status (str): 'success' or 'error'.
    msg (str): Human-readable execution summary.
    tool_name (str): Tool identifier 'chroma_symmetry'.
    output_dir (str): Run-specific output directory under tool_result/chroma_toolkit_result.
    output_file (str): Target output structure file path.
    group (str): Effective symmetry group used in generation.
    length (int): Effective residue length per chain.
    steps (int): Effective sampling step count.
    num_chain_neighbors (int): Effective neighbor-chain parameter.
    langevin_factor (float): Effective Langevin factor.
    inverse_temperature (float): Effective inverse temperature.
    device (str): Effective execution device.
    format (str): Effective output format.

How to use tool chroma_symmetry :

response = await client.session.call_tool(
    "chroma_symmetry",
    arguments={
        "group": "C_3",
        "length": 80,
        "steps": 500,
        "num_chain_neighbors": 2,
        "langevin_factor": 8.0,
        "inverse_temperature": 8.0,
        "device": "cuda:0",
        "fmt": "cif",
        "dry_run": False
    }
)
result = client.parse_result(response)
output_file = result["output_file"]

Example parameter sets

# 1) Main mode
{
    "group": "C_3",
    "length": 80,
    "steps": 500,
    "num_chain_neighbors": 2,
    "langevin_factor": 8.0,
    "inverse_temperature": 8.0,
    "device": "cuda:0",
    "fmt": "cif",
    "dry_run": False
}

# 2) Variant mode
{
    "group": "D_2",
    "length": 100,
    "steps": 50,
    "num_chain_neighbors": 3,
    "langevin_factor": 6.0,
    "inverse_temperature": 10.0,
    "fmt": "pdb",
    "dry_run": True
}

---

⚠ Mandatory Output File Download (L3 Principle 14)

After calling this tool, you MUST download all output structure files from the MCP server to the local workspace using server_file_to_base64. A tool call is NOT considered complete until its output files have been downloaded and verified locally (ls -la <file> — size must be > 0).

import base64, os
response = await client.session.call_tool(
    "server_file_to_base64",
    arguments={"file_path": result["output_file"]}  # or relevant output field
)
dl = client.parse_result(response)
local_path = "stepNN_descriptive_name.ext"
with open(local_path, "wb") as f:
    f.write(base64.b64decode(dl["base64_string"]))
assert os.path.getsize(local_path) > 0, f"Download failed: {local_path}"

Download policy: All structure output files are Category A (user-critical) — essential for user verification, downstream analysis, and reproducibility. When in doubt, download. Over-collection is always preferred over under-collection.

⚠ Mandatory Design Count Verification (L3 Principle 11)

After scaffold generation, verify the actual number of scaffolds produced matches the requested num_samples. Report the ACTUAL count.