PatSight-MolVision

PatSight MolVision molecular image recognition: calls Patent Extractor API on 1-N structure images, returns SMILES/SDF and compute properties by rdkit. First...

This skill uploads user images to an external PatSight Patent Extractor API and requires PatSight credentials (or an OPS token). Before installing: 1) Confirm you trust https://patent.xinsight-ai.com and the OPS token service at https://xops.xtalpi.com (these two domains are both used by the script). 2) Prefer using a short-lived PATSIGHT_TOKEN instead of storing your account/password in environment variables or config; verify where tokens are created/stored. 3) Do not send images that contain confidential or proprietary information unless the external service's privacy/retention policy is acceptable. 4) Note minor docs issues (example JSON in SKILL.md has a stray missing quote, and additional env vars like PATSIGHT_TOKEN and PATSIGHT_BASE_URL are used but not listed as required) — these are operational/quality concerns, not evidence of malice. If you need higher assurance, request the service's privacy/security documentation or run the script in an isolated environment and inspect network traffic while authenticating.

Type: OpenClaw Skill Name: ocsr Version: 1.0.1 The 'ocsr' skill bundle provides molecular image recognition by interfacing with the PatSight Patent Extractor API. The primary script, `scripts/run_ocsr.py`, handles image validation, authentication via account credentials or tokens, and uses RDKit to compute molecular properties. All network activity is directed to legitimate service endpoints (xinsight-ai.com and xtalpi.com), and the code follows standard practices for API interaction and report generation without any signs of malicious intent, data exfiltration, or prompt injection.