byted-marketing-agent-inspiration-insight

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly asks the user for AccessKey/SecretKey and instructs storing them in shell variables and appending them as --ak/--sk command-line arguments for script calls, which requires the agent to handle and potentially output secrets verbatim and exposes them on the CLI.

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

• Third-party content exposure detected (high risk: 0.90). Yes — the skill's required workflow (SKILL.md Step 1–3) and scripts/openapi_client.py explicitly call a Volcengine OpenAPI (cdp-saas.cn-beijing.volcengineapi.com) to list industries and query "hot_video" records that contain user-generated content (抖音链接, ASR/OCR text, keyframe URLs, analysis/replication prompts), which the agent must read and interpret to drive sorting, detail expansion, and generation actions — creating a clear avenue for indirect prompt injection.