ckb-ccc-playground
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill describes an online sandbox environment (
live.ckbccc.com) designed specifically to execute TypeScript code in the browser. This is the intended primary purpose of the documented tool. - [REMOTE_CODE_EXECUTION]: The documentation explains a feature where code can be loaded directly from an external URL via the
?src=query parameter. The skill provides clear security mitigation advice, instructing users to manually inspect the source before clicking 'Run' and warning against running unreviewed code while connected to a mainnet signer. - [INDIRECT_PROMPT_INJECTION]: The ability to load and execute code from external sources (Nostr relays or raw file URLs) creates a vulnerability surface where untrusted data enters the execution context.
- Ingestion points: Code loaded via the
?src=parameter or shared via Nostr links inSKILL.md. - Boundary markers: The skill explicitly includes warnings to inspect code before execution and to verify network settings.
- Capability inventory: The environment provides a pre-connected blockchain signer and transaction visualization capabilities.
- Sanitization: The skill relies on manual human review as the primary sanitization method before the code is executed.
Audit Metadata