same-outlet-ranker
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, creating a potential surface for indirect prompt injection.
- Ingestion points: Processes user-provided 'pitch' text and external data retrieved from URLs (profile pages, byline articles, and social media) as described in the retrieval section of SKILL.md.
- Boundary markers: The instructions explicitly state that the platform's 'ethical floor' and anti-spam doctrine (e.g., skills/ETHICS.md) take precedence over local instructions if a conflict occurs.
- Capability inventory: The skill utilizes web search capabilities, the medialyst database, and interacts with other agent skills like journalist-fit-check and find-journalists to perform its function.
- Sanitization: Includes 'Hard Gates' to detect 'slop tells' in the input pitch, which triggers a stop in execution and hands the process off to a 'meanest-editor' skill for sanitization and review.
Audit Metadata