story-origin-check
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection because it ingests and processes untrusted data from external websites.
- Ingestion points: The skill uses
WebFetchto retrieve content from supplied evidence URLs and uses thenews-searchtool to collect titles, snippets, and metadata from search results (SKILL.md, Process steps 1, 3, and 4). - Boundary markers: The instructions do not explicitly require the use of boundary markers (such as XML tags or unique delimiters) or "ignore embedded instructions" warnings when the agent interpolates retrieved web content into its context.
- Capability inventory: The skill is authorized to perform network operations (web search and fetching) and write structured analysis results to the local filesystem in the
origin_findings.jsonfile. - Sanitization: There are no specific instructions for sanitizing, filtering, or escaping the text retrieved from external news articles before the LLM processes it, which could allow a malicious article to influence the agent's logic.
Audit Metadata