iblai-api-studio-course-create
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill instructions require reading sensitive session identifiers and CSRF tokens from a local environment file (
studio.env). Accessing files with.envextensions is flagged as a sensitive path access. - [COMMAND_EXECUTION]: The skill uses shell commands, specifically
curlfor network operations andpython3 -cfor parsing JSON responses. The Python code is a static snippet used to extract a specific field from the API response. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, such as course summaries and user roles retrieved from the Studio API. This creates a potential surface for indirect prompt injection if the external data contains malicious instructions intended to influence the agent's behavior.
- Ingestion points: External data is ingested via GET requests to
/api/contentstore/v2/home/courses,/api/ibl/manage/course/, and/api/ibl/users/manage/roles/. - Boundary markers: None explicitly defined in the prompt instructions to isolate external data.
- Capability inventory: The skill has capabilities for network operations (
curl), file reading (studio.env), and command execution (python3). - Sanitization: No specific sanitization or filtering of the API-returned text is mentioned before it is processed by the agent.
Audit Metadata