industry-trends-weekly
Fail
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: CRITICALPROMPT_INJECTIONSAFE
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its reliance on external news sources.
- Ingestion points: The 'Workflow' involves gathering news and trends from the live web using web search tools.
- Boundary markers: The instructions do not define delimiters or provide 'ignore instructions' warnings for the data gathered from the web, which could allow malicious content in news articles to influence agent behavior.
- Capability inventory: The skill utilizes web search and optional read-only access to HubSpot data, but lacks the ability to execute system commands or write to the file system.
- Sanitization: There are no instructions for sanitizing or validating the content retrieved from external URLs.
- [SAFE]: No executable scripts, binaries, or configuration files are present in the skill. The logic is entirely expressed in natural language markdown.
- [PROMPT_INJECTION]: The metadata field 'author' contains a domain (Builder.org) that has been flagged by automated security scanners. While this domain is not used for code downloads or active execution within this skill, its presence in the metadata warrants caution.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata