nature-machine-intelligence
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is entirely comprised of markdown text providing guidance on journal fit and submission standards. There are no scripts, binaries, or automated shell commands included.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing research manuscripts (untrusted data ingestion). However, it does not utilize any exploitable capabilities like file-writing, network requests, or system command execution, rendering the injection surface safe.
- [EXTERNAL_DOWNLOADS]: Instructions are provided to manually re-check official Springer Nature sites for the latest guidelines. These are informational references for the agent's retrieval capabilities and do not represent malicious remote code execution or automated downloads of untrusted content.
Audit Metadata