Relative binding free-energy and activity-label prediction workflows using PBCNet 2.0 on SciMiner, with Gnina docking and PDB/database retrieval to complete...
Security Analysis
medium confidenceThe skill’s scientific workflow is disclosed, but its bundled PDB helper exposes a broad generic HTTP client and optional arbitrary raw-response file writes that exceed the narrow PDB-retrieval purpose.
The main FEP/PBCNet workflow, SciMiner API-key use, docking, and PDB retrieval are coherent with the stated purpose, but the bundled rest_request.py accepts caller-chosen destinations, headers, query parameters, and request bodies rather than enforcing the documented RCSB-only endpoints.
The RCSB sub-skill instructions name three expected RCSB base URLs, but the executable script does not enforce that scope and also accepts absolute http/https paths, making the runtime authority broader than the skill description.
The package contains Markdown skill instructions and one Python helper script; there is no evidence of hidden installer behavior, package installation, obfuscation, or automatic startup.
Network access to SciMiner and RCSB is expected, and the SciMiner credential file is disclosed, but a generic POST-capable HTTP helper with arbitrary headers is not well contained for a narrowly scoped scientific retrieval helper.
The script can save raw responses only when save_raw is enabled, but raw_output_path is caller-controlled and can write outside a controlled skill directory wherever the process has permission.
Guidance
Install only if you are comfortable with the agent reading a SciMiner API key from ~/.config/sciminer/credentials.json, uploading molecular/receptor data to SciMiner, and running a bundled Python HTTP helper. Keep use limited to the documented SciMiner and RCSB endpoints, avoid providing sensitive headers or payloads to the PDB helper, and use raw response saving only to a path you explicitly choose.
Latest Release
v1.0.0
- Initial release of FEP-alternative skill for lead optimization using SciMiner's PBCNet 2.0. - Automates prediction of relative binding free energies or activity labels with robust file-completion workflows using Gnina docking and PDB/database retrieval. - Includes flexible workflow selection: direct PBCNet invocation, crystal-guided docking, or PDB retrieval with ligand similarity filtering. - Ensures rigorous validation of required inputs, stepwise file preparation, and strict adherence to official SciMiner tool documentation. - User credential handling via external SciMiner API key; precise error handling for missing credentials. - Provides clear user guidance and enforced API doc compliance for every SciMiner tool invocation.
More by @sciminer
Published by @sciminer on ClawHub
