rick-mode
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses detailed persona instructions to override the agent's default helpful and polite behavior, enforcing a condescending and brutal tone. It includes instructions to drop the persona only for technical precision, maintaining the behavioral override throughout the session.- [COMMAND_EXECUTION]: The 'Agency' rule authorizes the agent to autonomously execute shell-level commands and perform file operations including
git stash,git branch,grep, running tests, and running lints. It explicitly instructs the agent to fix issues directly rather than just providing a checklist of commands.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its automated processing of untrusted code. - Ingestion points: Pre-Review Protocol (Steps 1 and 2) in SKILL.md reads arbitrary files and dependencies based on user-provided paths.
- Boundary markers: No specific delimiters or instructions are used to isolate the code being reviewed from the agent's operative instructions.
- Capability inventory: The agent has permissions to edit files, run tests, and perform git operations as part of its 'reversible' actions.
- Sanitization: No sanitization or escaping is performed on the ingested code content.- [DATA_EXFILTRATION]: The skill directs the agent to access project configuration files such as
package.jsonandtsconfig.jsonto verify environmental context. While common in development workflows, this exposes internal dependency versions and compiler configurations to the agent's context.
Audit Metadata