grokipedia-cli
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
grokipediaCLI tool to perform various operations, including article retrieval, full-text searching, and exporting data to the local filesystem (e.g.,grokipedia backup --all --output grokipedia-backup.zip).- [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection or command injection by directly interpolating user-provided text into the--queryflag of thegrokipedia searchcommand. Evidence Chain: 1. Ingestion points: User-controlled search queries are processed bySKILL.md. 2. Boundary markers: No delimiters or instructions are provided to the agent to treat the input as literal data. 3. Capability inventory: The agent executes shell commands using thegrokipediautility. 4. Sanitization: No input validation or escaping mechanisms are specified for the interpolated query strings.
Audit Metadata