grammarly-common-errors
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute
curlcommands for verifying connectivity to Grammarly's API ecosystem. These operations are limited to a well-known service domain and use environment variables for authentication, following standard security practices. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process diagnostic information and API responses which could potentially contain untrusted content. It mitigates this risk by providing explicit instructions to use synthetic test strings and correlation IDs, and strictly forbids the retention or reporting of actual user text, tokens, or identities.
Audit Metadata