okx-ai-builder-integration
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches reference implementation demos and technical documentation from the author's official GitHub repositories (
github.com/okx/builder-integration-demoandgithub.com/okx/ai-builder-openapi-md). - [COMMAND_EXECUTION]: Provides instructions for installing the official OKX Trade CLI utility (
@okx_ai/okx-trade-cli) and cloning demo projects via Git. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external documentation and code from remote repositories to guide the agent. While this creates a data processing surface, the source is limited to official vendor-controlled infrastructure.
- Ingestion points: The agent is instructed in
SKILL.mdto read README, guides, and code samples from theokx/builder-integration-demorepository. - Boundary markers: Explicitly absent; the skill instructs the agent to trust the repository as the authoritative source for routing and implementation details.
- Capability inventory:
SKILL.mddescribes capabilities including shell command execution for environment setup and repository cloning. - Sanitization: Absent; no specific sanitization of the remote repository content is performed beyond recommending user review of the adapted code.
Audit Metadata