rocketsim-setup
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses local shell commands including
pgrep,ps,mdfind, andlsto identify the installation path of the RocketSim application and verify the presence of its CLI and bundled skill files. - [INDIRECT_PROMPT_INJECTION]: The agent is instructed to read and follow instructions from a
SKILL.mdfile found within the RocketSim application bundle on the local filesystem. This ensures the agent uses guidance that matches the installed version of the tool. - Ingestion points: Reads instructions from the application's resources directory at
$APP_PATH/Contents/Resources/Agent-Skill/*/SKILL.md. - Boundary markers: None; the agent is directed to follow the instructions found in the external file for the duration of the session.
- Capability inventory: The skill uses process management and filesystem discovery utilities to locate the target application.
- Sanitization: Discovery is validated against a specific bundle identifier (
com.swiftLee.RocketSim) associated with the vendor.
Audit Metadata