blog-brief
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites via web search in Steps 2, 3, and 4 of the workflow to perform competitive analysis and gather statistics. This introduces a vulnerability where malicious instructions embedded in third-party websites could potentially influence the agent's output. \n
- Ingestion points: External content fetched via WebSearch in
SKILL.md(Step 2: Keyword Research, Step 3: Competitive Analysis, and Step 4: Statistics Research). \n - Boundary markers: Absent; the instructions do not specify any delimiters or safety prompts to isolate external data. \n
- Capability inventory: The skill utilizes web search capabilities and performs file writes (
briefs/[slug]-brief.md) as described inSKILL.md. \n - Sanitization: Absent; there is no explicit instruction to sanitize or filter instructions from the retrieved web content. \n- [EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to perform web searches and visit external websites to retrieve SEO data, competitor heading structures, and industry statistics from various sources including Pixabay and Unsplash.
Audit Metadata