byted-bp-searchinfinity-web-search

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly asks the user to paste their BytePlus API key into the chat and instructs the agent to accept, trim, overwrite, and use it (including running verification via a --api-key flag), which requires the model to receive and potentially include the secret verbatim, creating an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.85). The skill calls BytePlus Searchinfinity at runtime and then ingests the API’s returned WebResults fields (e.g., Title, Summary/Snippet, Url) into the agent’s context; those fields originate from arbitrary public web pages and can contain outsider-authored free text (indirect prompt injection risk).