viral-post-writer
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external content from the internet that could be manipulated to bias or hijack the agent's output instructions.\n
- Ingestion points: The workflow requires the agent to perform
web_searchandweb_browseto research product details, competitor comparisons, and user testimonials (SKILL.md Step 2 and 2.5).\n - Boundary markers: The instructions do not define delimiters or provide specific prompts for the agent to distinguish between valid research data and potentially malicious instructions embedded within retrieved web pages.\n
- Capability inventory: The skill is configured to use
web_searchandweb_browsetools (agents/openai.yaml). It generates formatted text for social media and does not have permissions for file writes, shell execution, or unauthorized network exfiltration.\n - Sanitization: There are no specified mechanisms for filtering or sanitizing content retrieved from external sources before it is used for post generation.
Audit Metadata