content-strategy
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is composed of purely instructional content intended to guide an AI agent in the role of a content strategist. It does not include any scripts, executables, or commands that could compromise system security.
- [PROMPT_INJECTION]: The skill defines a process for analyzing external data, including web search results from forums (Reddit, Quora), customer call transcripts, and SEO data exports. This ingestion of external data is a known surface for indirect prompt injection, but it is necessary for the skill's primary function of audience and keyword research.
- Ingestion points: Keyword data exports, sales call transcripts, survey responses, and forum search results (SKILL.md).
- Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are provided for these data sources.
- Capability inventory: The skill relies on the agent's file-reading and web-search tools.
- Sanitization: The instructions do not specify any validation or sanitization steps for the external content before analysis.
Audit Metadata