alapi

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs constructing and issuing commands like python3 "$SCRIPT" --json call {path} --token {用户的token} and prioritizes an explicit --token over environment variables, which requires the LLM to insert the user's secret token verbatim into generated commands/requests (an exfiltration risk) despite advising not to echo tokens in final replies.

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

• Third-party content exposure detected (high risk: 1.00). The skill's CLI and workflow explicitly fetch and parse public ALAPI web content (e.g., scripts/alapi.py calls https://v3.alapi.cn/frontend/api/* and https://v3.alapi.cn/openapi/{id}.json, and SKILL.md/README instruct use of the documentation URL https://www.alapi.cn/api/{id}/introduction), meaning the agent reads untrusted third-party pages/OpenAPI specs as part of its decision and action flow, so third-party content can influence tool use.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime CLI (scripts/alapi.py) performs live requests to https://v3.alapi.cn/openapi/{id}.json (and related https://v3.alapi.cn frontend/api endpoints) to fetch OpenAPI specs and search results which are ingested at runtime and directly drive prompt-driven code generation, so this external URL is a required runtime dependency that can control agent outputs.