drafts
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to execute a local wrapper script located atscripts/api-request.shfor all communications with the Spotify Ads API. - [DYNAMIC_EXECUTION]: Within the
Setupsection ofSKILL.md, the instructions direct the agent to useeval $(api --env)to dynamically assign environment variables such as tokens and account IDs from the output of a local script. - [INDIRECT_PROMPT_INJECTION]: The skill's
buildoperation ingests unstructured user descriptions to generate campaign hierarchies, which could be exploited to deliver malicious instructions. - Ingestion points: User-supplied plain-text campaign descriptions processed by the
buildcommand defined inSKILL.md. - Boundary markers: There are no explicit instructions or delimiters defined to ensure the agent ignores or sanitizes instructions embedded within the user's description.
- Capability inventory: The skill maintains capabilities to execute shell commands via
Bash(specifically for API requests) and read local filesystem content via theReadtool. - Sanitization: The skill does not describe any specific validation or escaping mechanisms for the natural language input before it is used to determine API request parameters.
Audit Metadata