lov-thesis-polish
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes MBA thesis content provided by users, which represents an ingestion point for untrusted data. Malicious instructions or patterns embedded within this content could potentially influence the agent's behavior during the polishing or evaluation process.
- Ingestion points: The agent is instructed to read markdown thesis text or files provided by the user, as specified in
SKILL.md(Step 1). - Boundary markers: The instructions do not define clear delimiters or specific guidance to ignore instructions that might be present within the processed thesis data, such as HTML comments or hidden markdown text.
- Capability inventory: The skill utilizes file system tools (
ReadandWrite) to manage the thesis documents and an interaction tool (AskUserQuestion) for strategy confirmation. - Sanitization: There are no explicit instructions for sanitizing, filtering, or validating the input text before it is analyzed and modified by the agent.
Audit Metadata