onyx-cli
Fail
Audited by Gen Agent Trust Hub on Apr 15, 2026
Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's setup instructions include a command to move the compiled binary using administrative privileges:
sudo mv onyx-cli /usr/local/bin/. This represents a privilege escalation vector involving system-level file modification. - [EXTERNAL_DOWNLOADS]: The skill provides instructions to download the
onyx-clitool via the Python Package Index and to build from source using the vendor's GitHub repository atgithub.com/onyx-dot-app/onyx/cli. These resources are provided by the official vendor. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes content from external company sources.
- Ingestion points: Results from the
onyx-cli askcommand which pulls data from connected sources like Confluence, Slack, and Google Drive. - Boundary markers: The skill does not define specific delimiters or instructions to prevent the agent from being influenced by instructions embedded within the retrieved document content.
- Capability inventory: The agent is instructed on how to execute CLI commands using the
onyx-clitool. - Sanitization: There is no evidence of sanitization or filtering applied to the retrieved knowledge base data before it is presented to the agent.
Recommendations
- AI detected serious security threats
Audit Metadata