agkan-icebox
Pass
Audited by Gen Agent Trust Hub on Apr 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
agkanCLI to interact with task data, executing commands to list and retrieve specific task information.\n- [PROMPT_INJECTION]: Analysis identifies an indirect prompt injection surface where external task content is incorporated into agent prompts.\n - Ingestion points: Task titles and bodies are fetched via
agkan task listandagkan task getas defined in SKILL.md.\n - Boundary markers: Absent. Task content is placed directly into sub-agent prompts without delimiters or instructions to ignore instructions found within the data.\n
- Capability inventory: The skill environment allows shell command execution via
agkan, and the sub-agent has the ability to read and follow procedures from local skill files.\n - Sanitization: The skill does not perform validation or escaping of the task data before it is processed by the AI models.
Audit Metadata