skills/okx/agent-skills/okx-cex-earn/Gen Agent Trust Hub

okx-cex-earn

Pass

Audited by Gen Agent Trust Hub on Nov 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the official vendor package '@okx_ai/okx-trade-cli' via npm during the setup process. This is a functional requirement for the skill to operate.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands via the 'okx' CLI to fetch account data and perform financial operations such as subscribing to or redeeming from interest-bearing products.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests data from external API sources and has high-impact capabilities.
  • Ingestion points: API responses from market tickers, earn product lists, and user account balances (e.g., 'okx earn savings balance', 'okx earn dcd products').
  • Boundary markers: While the skill requires explicit user confirmation for write operations, it lacks specific instructions to the agent to treat raw tool outputs as untrusted data.
  • Capability inventory: The skill can perform high-privilege actions including asset purchases, redemptions, and internal account transfers (documented in 'references/workflows.md').
  • Sanitization: The agent is instructed to use the '--json' flag for all queries and render results in structured Markdown tables to normalize the output.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 13, 2026, 02:26 PM