zephyr
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use Bash at runtime to check the existence and value of the
ZEPHYR_BAKE_LOG_DIRenvironment variable for directory resolution.- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill reads and processes bake log files (Markdown) to extract bake events and update YAML blocks. * Ingestion points: Bake log files (.md) located in theZEPHYR_BAKE_LOG_DIR. * Boundary markers: None; the skill relies on string-based sentinel patterns ('For Zephyr N') to identify data. * Capability inventory: Shell command execution (Bash) and local file write operations. * Sanitization: There is no instruction to sanitize or validate the content of the bake log entries before processing or interpolating them.
Audit Metadata