9router-embeddings
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: References setup instructions from a remote repository at
https://raw.githubusercontent.com/decolua/9router/refs/heads/master/skills/9router/SKILL.md.- [COMMAND_EXECUTION]: Provides templatecurlcommands for interacting with the 9Router API endpoints, using user-defined environment variables for the URL and authentication key.- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text to generate embeddings, which creates a surface for potential indirect instructions if the input data is sourced from untrusted external content. - Ingestion points: The
inputfield in the POST request body withinSKILL.md. - Boundary markers: None identified.
- Capability inventory: Network communication using
curlandfetchto external endpoints. - Sanitization: No explicit sanitization of input strings before being sent to the embedding provider.
Audit Metadata