nature-data
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves exclusively as an instructional guide and template framework for constructing academic Data Availability statements. It contains no executable scripts, binaries, tool invocations, or network capability configurations.
- [INDIRECT_PROMPT_INJECTION]: Assessment of the untrusted data ingestion surface:
- Ingestion points: Untrusted inputs enter the context when users provide manuscript text, drafts, or data descriptions as described in
SKILL.mdandstatic/core/workflow.md. - Boundary markers: Absent; there are no instructions to encapsulate user content using delimiters or explicit containment rules.
- Capability inventory: None; the skill lacks any external tools, shell execution capabilities, file system write operations, or network communication facilities.
- Sanitization: Absent; input validation or escaping mechanisms are not specified for user-supplied descriptions.
Audit Metadata