update-audit
Warn
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill defines a 'YOLO mode' that instructs the agent to carry out actions, such as updates or tool replacements, without seeking repeated approval after the initial mode choice. This reduces human oversight for administrative operations on the host system.
- [INDIRECT_PROMPT_INJECTION]: The skill's audit process ingests untrusted metadata from the host environment, which could contain malicious instructions. 1. Ingestion points: OS status, app versions, driver details, and package manager health (SKILL.md). 2. Boundary markers: No delimiters or specific instructions to ignore embedded commands are present in the provided files. 3. Capability inventory: The skill is designed to perform system modifications including updates and tool replacements. 4. Sanitization: No sanitization or validation of the retrieved system metadata is described.
- [EXTERNAL_DOWNLOADS]: The primary purpose of the skill is to identify and perform updates, which inherently requires fetching and executing software installers or code from external repositories.
Audit Metadata