reference-style-sync
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external bibliographic data files, which represents an ingestion surface for indirect prompt injection attacks.
- Ingestion points: Bibliographic data is loaded from user-specified file paths in
scripts/main.pyvia theReferenceSync.loadmethod and parsed through various format-specific methods (e.g.,parse_bibtex,parse_ris). - Boundary markers: The skill instructions and scripts do not implement explicit boundary markers or sanitization logic to ignore instruction-like text embedded within bibliographic fields (e.g., title, abstract).
- Capability inventory: The skill has the capability to write processed data to the local file system via the
ReferenceSync.exportmethod inscripts/main.pyand print output to the console. - Sanitization: While the script uses regular expressions and standard parsers to restructure metadata, it does not filter for potential malicious instructions in the content.
Audit Metadata