content-ideas
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill collects untrusted data from users via a structured intake form including fields like 'Client name', 'Target audience', and 'Primary goal'. This data is then used to generate content ideas, creating a potential surface for instructions embedded within these fields to influence the agent's behavior.\n
- Ingestion points: Found in
SKILL.mdunder the 'Required Input' section.\n - Boundary markers: Absent. There are no instructions to use delimiters or ignore instructions found within the provided data.\n
- Capability inventory: The skill is limited to generating Markdown text output. It does not execute code, perform network operations, or access the filesystem.\n
- Sanitization: Absent. The skill does not define any validation or filtering logic for the input fields.
Audit Metadata