j-patent-strategy-analyzer
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits a vulnerability surface for indirect prompt injection as it ingests and processes untrusted external data from two primary sources: \n
- Ingestion points: Uploaded Excel patent lists (referenced in
references/data_prep.md) and web search results for company information and financials (defined inSKILL.md). \n - Boundary markers: There are no explicit instructions to use delimiters or 'ignore embedded instructions' warnings when processing these inputs before they enter the agent's context. \n
- Capability inventory: The agent has the capability to install Python packages (
runtime.apply_sync), read local files (pd.read_excel), execute arbitrary Python code for data processing and chart generation, and perform web searches. \n - Sanitization: The skill lacks explicit requirements for sanitizing or escaping the external content before it is interpolated into the HTML template (
references/html_skeleton.md) and specific report sections (references/section_specs.md).
Audit Metadata