twitter-thread-creation
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation references external installation instructions for the
beltCLI hosted on a public GitHub repository (inference-sh/skills). This is provided to the user as a prerequisite for using the skill's functionality. - [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests data from external research tools and uses it to perform automated actions, creating a potential surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent's context through the
tavily/search-assistanttool, which is used to research statistics and source material. - Capability inventory: The skill utilizes the
x/post-createtool to publish content to social media andinfsh/html-to-imagefor visual content generation. - Boundary markers: The provided command templates do not include specific delimiters or "ignore" instructions to separate the ingested research data from the agent's primary task instructions.
- Sanitization: There is no evidence of content validation or sanitization of the research output before it is interpolated into the social media post templates.
Audit Metadata