fb-post-writer
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied topics, angles, and rough drafts which are then interpolated into post drafts and sent to publishing tools. This creates a surface for indirect prompt injection.\n
- Ingestion points: User-provided content (topic, angle, rough draft) gathered in SKILL.md Step 1, and the voice-profile.md reference file.\n
- Boundary markers: The skill lacks explicit delimiters or instructions for the agent to ignore embedded commands within user input, although it does include a mandatory human approval step before publishing in SKILL.md Step 7.\n
- Capability inventory: The skill utilizes lib.publish (SKILL.md Step 8) to send content to social media platforms and lib.illustrate (SKILL.md) to generate images via external services.\n
- Sanitization: There are no mentioned processes for sanitizing or escaping user-provided input before it is used to generate the final draft or passed to the publishing tool.
Audit Metadata