website-copy-reframe
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external sources, which creates a surface for indirect prompt injection attacks. \n
- Ingestion points: In
SKILL.md(Step 1), the agent is directed to scrape content from live URLs or ingest user-provided drafts. \n - Boundary markers: There are no instructions in the skill files to use specific delimiters or to explicitly warn the model to ignore instructions contained within the ingested website copy. \n
- Capability inventory: The skill possesses the capability to write local files in
.mdor.docxformats and can interact with external services like Google Drive and Google Workspace (as described in the Output contract). \n - Sanitization: The skill lacks defined sanitization or filtering steps for the ingested external content before it is used to generate analysis and new marketing copy.
Audit Metadata