The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides examples using npx onekey, which downloads and executes a package from the public npm registry. This package is not from a well-known or trusted organization.
[COMMAND_EXECUTION]: Provides instructions for executing shell commands including curl for API interaction and npx for CLI tool execution, which involves running external code.
[DATA_EXFILTRATION]: Transmits user-supplied data and environment variables (DEEPNLP_ONEKEY_ROUTER_ACCESS) to external domains agent.deepnlp.org and craftsman-agent.aiagenta2z.com. While this is functional for the service, it represents a network data flow to non-whitelisted domains.
[PROMPT_INJECTION]: The skill processes untrusted user input via the prompt field which is interpolated into API requests, creating a surface for indirect prompt injection where malicious instructions in the prompt could influence the agent's behavior.
Ingestion points: The prompt and brief_inputs fields in SKILL.md (used in API calls and CLI commands).
Boundary markers: None identified in the prompt interpolation process.
Capability inventory: Network requests via curl, package execution via npx.
Sanitization: No evidence of input sanitization or validation to prevent prompt injection attacks.

icon-designer-generator