agent-sdk-builder

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill explicitly instructs the agent to fetch and research public content from docs.openhands.dev/llms.txt and to clone example and docs repositories from GitHub (OpenHands/software-agent-sdk and OpenHands/docs), which are untrusted third-party sources the agent must read and that can materially influence implementation decisions and tool use.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent at runtime to fetch and clone external resources (https://docs.openhands.dev/llms.txt, https://github.com/OpenHands/software-agent-sdk/tree/main/examples/01_standalone_sdk, and https://github.com/OpenHands/docs/tree/main/sdk), which will be loaded into the workspace and used to shape prompts/implementation and may include executable example code, so the fetched content can directly control prompts or code the agent relies on.