scientific-writing-outline
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions define a legitimate workflow for scientific writing and do not contain any malicious patterns such as prompt injection, obfuscation, or unauthorized data exfiltration. The instructions focus on accuracy and preventing hallucinations in a scientific context.
- [DATA_EXFILTRATION]: The skill is designed to read from local files such as
story.mdand repository artifacts (e.g., CSVs, scripts, notebooks) and access the web for citation verification. These actions are strictly aligned with the intended purpose of scientific manuscript preparation and do not involve access to sensitive system or credential files. - [PROMPT_INJECTION]: The skill ingests data from
story.mdand repository files. While this represents a theoretical attack surface for indirect prompt injection, the risk is mitigated as the instructions guide the agent toward rigorous validation of the content against a thesis and existing literature rather than execution of instructions within that data.
Audit Metadata