Protein-Ligand Docking

Run a protein-ligand docking workflow for research questions about target binding, selectivity, and structural plausibility. Use this skill when the user ask...

This skill appears to implement a legitimate protein–ligand docking pipeline, but before installing or running it you should: (1) Confirm that the environment has the required external tools (OpenBabel 'obabel', AutoDock Vina) and Python packages (Biopython, NumPy, RDKit, python-docx, etc.). The registry metadata currently omits these requirements, so the skill may fail or produce misleading results if they are missing. (2) Verify the missing referenced file: SKILL.md points to references/alphafold_multimer_colab.ipynb but that file isn't in the bundle—ask the publisher for the notebook or update the workflow to make explicit how to obtain it. (3) Review the scripts locally before running: they invoke subprocesses (obabel, vina) and read/write files in the working directory; ensure you run them in an isolated workspace with input files you trust. (4) If you plan to allow autonomous agent invocation, restrict network access or review logs—while this skill doesn't request secrets, autonomous execution plus external web access can increase risk if you haven't vetted the inputs and outputs. (5) Consider running the scripts manually in a controlled environment first to confirm outputs and to ensure the correct binary paths and versions are used. If you want, I can produce a checklist of the exact packages, command-line tools and minimum versions to install, or highlight the precise JSON/field mismatches to fix in the code.

Type: OpenClaw Skill Name: protein-ligand-docking Version: 2.3.0 The skill bundle provides a legitimate and well-structured workflow for protein-ligand docking and bioinformatics analysis. The Python scripts (such as step6_vina_docking.py and step3_alignment.py) use standard libraries like Biopython and NumPy to process scientific data and invoke external tools (AutoDock Vina, OpenBabel) via safe subprocess calls. There is no evidence of malicious intent, data exfiltration, or prompt injection; the instructions in SKILL.md and the decision guide are strictly aligned with the stated research purpose.