content-to-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze untrusted external data, such as competitor social media posts and audience engagement metrics, which could theoretically contain malicious instructions.
- Ingestion points: Sections 2 (Distribution Reverse Engineering) and 3 (Multi-Platform Content Repurposing) involve exporting and analyzing external content like the "top 50 performing posts from the last 90 days" and "competitor content landscape."
- Boundary markers: Absent. There are no instructions for the agent to use delimiters or ignore embedded commands in the analyzed external content.
- Capability inventory: None. The skill file is purely markdown instructions and does not define tools, network calls, or file-system writes.
- Sanitization: Absent. No validation or filtering instructions are provided for the external data ingested.
Audit Metadata