ganjoor
Warn
Audited by Gen Agent Trust Hub on Nov 21, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation provides instructions to install the
ganjoor-mcppackage using theuvxtool. This involves downloading and executing third-party code from a package repository that is not associated with a verified trusted organization. - [INDIRECT_PROMPT_INJECTION]: The skill fetches and displays poetry, meanings, and user comments from
ganjoor.net, creating a surface for indirect instructions. - Ingestion points: Content is retrieved from the Ganjoor website through several tools including
gj_poemandgj_search. - Boundary markers: The instructions lack explicit delimiters or warnings to prevent the agent from following instructions that might be embedded within the retrieved poetic text or meanings.
- Capability inventory: The tools are described as read-only and limited to the retrieval and display of literary data.
- Sanitization: There is no evidence of sanitization or validation of the data retrieved from external sources before it is presented to the agent.
Audit Metadata