RxnIM
PassAudited by ClawScan on May 1, 2026.
Overview
This skill is a coherent chemistry image-parsing helper, but users should notice that uploaded images may be sent to HuggingFace and that optional local setup runs code from external repositories.
This appears suitable for its stated chemistry image-parsing purpose. Before installing or using it, decide whether your reaction images can be shared with the HuggingFace Space; for confidential chemistry data, prefer local deployment after reviewing the referenced GitHub repository and dependencies.
Findings (2)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Chemical reaction images may contain proprietary or unpublished research and would be shared with an external service when the online API path is used.
The workflow sends the user's reaction image, encoded as base64, to an external HuggingFace Space for processing.
POST https://CYF200127-RxnIM.hf.space/run/predict ... Body: { "data": ["data:image/png;base64,<base64_image>"] }Use the online API only for images you are comfortable sending to HuggingFace; use the documented local deployment path for confidential material.
If a user chooses local deployment, they would run externally sourced code and install dependencies on their machine.
The optional local setup relies on code, dependencies, and model weights from external sources that are not included in the reviewed artifact set.
git clone https://github.com/CYF2000127/RxnIM ... pip install -r requirements.txt ... 下载模型权重（~14GB） ... sh eval.sh
Review the referenced repository, requirements, and model download source before running the local deployment commands, preferably in an isolated environment.