The Agent Skills Directory

[SAFE]: The skill does not contain any malicious patterns. It is purely instructional and provides a workflow for scientific peer review without invoking any tools or external code.
[NO_CODE]: There are no scripts, binaries, or automated shell commands present in the skill files. All provided information is intended for the model's context to improve its qualitative analysis of research manuscripts.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to process untrusted external data (scientific manuscripts), it lacks any associated tools (e.g., network access, file system writes) that could be exploited via indirect injection. The risk is limited to text generation within the session.

peer-review