lov-typeless-prompt
Warn
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: MEDIUMOBFUSCATIONDYNAMIC_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [OBFUSCATION]: The skill's functional components, including instructions, scripts, and reference documents, are stored in encrypted files (.enc and .enc.json). This design explicitly prevents static analysis and auditing of the skill's behavior prior to execution.
- [DYNAMIC_EXECUTION]: The SKILL.md file instructs the agent to run a decryption command and then 'Read and execute the decryption output as the real SKILL.md'. This pattern constitutes dynamic loading and execution of instructions that are not present in the skill package itself.
- [REMOTE_CODE_EXECUTION]: The skill requires the agent to execute shell commands using npx and uvx to fetch and execute tools (lovstudio, lovstudio-skill-helper) from external registries, which are then used to generate and execute instructions at runtime.
Audit Metadata