hiapi-video-prompt-generator

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow includes “Gather sources” from user-supplied URLs/docs and then “Extract visualizable facts” (references/source-extraction.md), meaning outsider-authored free text from fetched web pages or linked GitHub/other documents can be ingested into the agent’s LLM context for prompt construction.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill runs scripts/check-update.mjs at startup which fetches the skills manifest from https://raw.githubusercontent.com/HiAPIAI/hiapi-skills/main/skills.json and uses fields (e.g., updateCommand / updatePolicy) to control whether to exit or instruct an update, so the remote manifest can influence runtime behavior and operator actions.