ielts-listening-review
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides and executes a local Python script (
scripts/ielts_buddy_api.py) to interact with the IELTS Buddy REST API. This script handles account binding and data retrieval tasks. - [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes external data from the IELTS Buddy API endpoints, which introduces a potential surface for indirect prompt injection if the external data contains malicious instructions.
- Ingestion points: Data enters the agent context through API calls such as
ielts_review_read_snapshotandielts_footprints_listfacilitated byscripts/ielts_buddy_api.pyfetching fromhttps://work.ieltsbuddy.igopx.cn/api/v1/agent. - Boundary markers: The instructions include guidelines for the agent to limit data output (e.g., "do not print the entire package to context") and to maintain conclusions of "incomplete" when data is truncated.
- Capability inventory: The skill can execute the
scripts/ielts_buddy_api.pyscript and make network calls to the vendor's API. - Sanitization: There is no explicit description of sanitization or input validation performed on the API responses before they are processed by the agent.
Audit Metadata