shift-handover-summarizer
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes Electronic Health Record (EHR) data through a JSON file input. This external data source could contain malicious instructions embedded within clinical notes or event descriptions.
- Ingestion points:
scripts/main.pyingests data from a file path provided via the--recordsCLI argument. - Boundary markers: The skill lacks explicit delimiters or instructions for the agent to ignore potentially malicious prompts contained within the records.
- Capability inventory: The skill possesses the ability to execute local Python scripts and perform file read/write operations.
- Sanitization: There is no evidence of text sanitization or filtering of the patient record content before it is processed by the agent.
- [INDIRECT_PROMPT_INJECTION]: The skill package includes a file named
shift-handover-summarizer_audit_result_v2.json. This file contains a self-referential security audit (Category 8e) claiming the skill is 'Production Ready' and has passed all security gates. This is a pattern used to bias or bypass automated security evaluations by presenting a pre-completed 'SAFE' verdict.
Audit Metadata