claude-code-dispatch
Warn
Audited by Gen Agent Trust Hub on Apr 19, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
bashtool to runclaudewith the--dangerously-skip-permissionsflag. This bypasses the security model of the CLI tool, allowing it to perform actions like file deletion or command execution without requiring human confirmation. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it interpolates task descriptions directly into the sub-agent's command-line prompt. An attacker could provide a task description containing malicious instructions that the sub-agent will execute autonomously due to the disabled safety checks.
- Ingestion points: The task description argument (
-p) in the CLI call. - Boundary markers: None; the input is not enclosed in delimiters that would signal the end of untrusted content.
- Capability inventory: File system access and command execution through the
bashtool. - Sanitization: The skill does not perform any escaping or verification of the task description before execution.
Audit Metadata