okx-cex-portfolio
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the
@okx_ai/okx-trade-clipackage from the official npm registry during the installation phase. This package is a scoped vendor resource belonging to the skill author (okx). - [COMMAND_EXECUTION]: The skill relies extensively on executing the
okxCLI tool to perform all read and write operations on the user's exchange account. This includes sensitive commands likeokx config showandokx account transfer. - [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes untrusted data from external sources and user inputs.
- Ingestion points: Data retrieved from OKX API endpoints (e.g., account balance, position history, billing records) and user-provided arguments for instrument IDs, currency filters, and profile names.
- Boundary markers: The instructions do not specify explicit delimiters for the agent to separate data from instructions during processing.
- Capability inventory: The skill has the capability to execute shell commands via the
okxCLI, perform network operations to interact with the OKX API, and read local configuration files (~/.okx/config.toml). - Sanitization: There are no documented sanitization or validation steps for the data retrieved from the API before it is presented to the agent context.
Audit Metadata