o-computer-personalization
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute the
o-computer-appearancecommand-line tool to inspect and change system appearance settings. - [INDIRECT_PROMPT_INJECTION]: The skill introduces a vulnerability surface by instructing the agent to "follow its instructions" from the output of the
o-computer-appearance --helpcommand. - Ingestion points: The agent is directed to process and obey output from a shell command (
o-computer-appearance --help). - Boundary markers: There are no instructions provided to treat the command output as data only or to ignore embedded instructions within that output.
- Capability inventory: The agent is authorized to execute the
o-computer-appearancetool. - Sanitization: There is no evidence of sanitization or filtering applied to the content returned by the command before the agent processes it.
Audit Metadata