glean-common-errors
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a diagnostic
curlcommand template to verify API connectivity. The command correctly uses an environment variable ($GLEAN_API_KEY) for authentication and targets the official service domain (glean.com), avoiding hardcoded credentials.- [INDIRECT_PROMPT_INJECTION]: The skill provides an error classification handler that processes data from external API responses (specified inSKILL.mdunderclassifyGleanError). This represents a standard surface for indirect prompt injection where untrusted data (error messages) is ingested into the agent's context. Boundary markers and sanitization are absent in the provided logic, but the skill's capabilities are limited toRead,Grep, andBashfor connectivity testing.
Audit Metadata