oled-intelligence-portal
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its ingestion of external data from news and patent search tools. This data is rendered into local HTML files without proper sanitization, creating a cross-site scripting (XSS) surface in the generated portal.
- Ingestion points: Data retrieved from
ls_news_vector_searchandmcp_patent-search__patsnap_searchas described in SKILL.md steps 4 and 5. - Boundary markers: No explicit delimiters or warnings are used to separate untrusted content from the portal templates.
- Capability inventory: The skill uses a Python script (
scripts/generate_portal.py) to perform file-write operations on the local system to generate the portal files. - Sanitization: The
generate_portal.pyscript uses Python f-strings to inject external data directly into HTML templates without performing escaping or validation, which could allow malicious scripts within the data to execute when the portal is opened in a browser. - [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/generate_portal.py) to process retrieved intelligence data and generate the multi-page HTML portal. The script uses standard libraries and processes data passed via environment variables or command-line arguments.
Audit Metadata