302ai-api-integration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly asks the user for their API key and instructs the agent to embed that exact key into generated code and headers (e.g., API_KEY = "{user_API_KEY}" and Authorization: Bearer {user_API_KEY}), which requires the LLM to handle and output 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: 1.00). The skill's required workflow (SKILL.md Steps 3 and 6) and the included script scripts/parse_api_list.py explicitly fetch public API listings from https://s.apifox.cn/apidoc/docs-site/4012774/llms.txt and instruct using WebFetch on user-selected API doc links (e.g., https://doc.302.ai/...), meaning the agent ingests open/public third‑party documentation which it parses and uses to drive API selection and code-generation decisions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly and obligatorily fetches remote API lists and docs at runtime (e.g., https://s.apifox.cn/apidoc/docs-site/4012774/llms.txt and user-selected https://doc.302.ai/*.md via WebFetch), and those fetched documents are parsed and injected into prompts/instruction-generation to drive code generation, so these URLs are runtime dependencies that directly control agent behavior.