ton-create-wallet
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions direct the agent to invoke system commands such as open, xdg-open, or start to automatically launch a web browser with the wallet dashboard URL. This is intended to streamline the user onboarding process.
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by taking output from the agentic_start_root_wallet_setup tool and passing it directly to a system shell command. 1. Ingestion points: The dashboardUrl value returned by the setup tool. 2. Boundary markers: No delimiters or boundary markers are defined for the tool output. 3. Capability inventory: The agent is authorized to execute shell commands to launch local applications. 4. Sanitization: No sanitization or validation of the URL is specified before it is passed to the opener commands.
- [SAFE]: All network references and domains, specifically agents.ton.org, are official resources belonging to the TON ecosystem and the skill's authoring organization.
Audit Metadata