line-openapi-skill

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

• Third-party content exposure detected (high risk: 0.70). The SKILL.md instructs the agent to call and parse responses from the public LINE Messaging API (https://api.line.me) — e.g., get:/v2/bot/profile, get:/v2/bot/channel/webhook/endpoint and message endpoints — which return user-generated/untrusted content that the agent is expected to read and could materially influence follow-up actions like replies or pushes.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill explicitly requires fetching the curated OpenAPI schema at runtime from https://raw.githubusercontent.com/holon-run/uxc/main/skills/line-openapi-skill/references/line-messaging.openapi.json (via the uxc link --schema-url command), and that fetched JSON directly determines the CLI/agent operation surface used to drive actions, so it is a runtime external dependency that can change agent behavior.