Screen Recording Online

edit screen recordings into polished screen recordings with this screen-recording-online skill. Works with MP4, MOV, WebM, AVI files up to 500MB. educators,...

This skill appears to do what it says: it uploads user video files to an external service (mega-api-prod.nemovideo.ai) for server-side editing and returns download links. Before installing, consider: (1) Privacy — any video you provide will be sent to that external API; do not upload sensitive recordings unless you trust the service and its policy. (2) Tokens — the skill uses NEMO_TOKEN (or will request an anonymous token) to authenticate; keep only tokens you are comfortable exposing to the agent. (3) Filesystem checks — the skill may check declared config paths and the install path for attribution headers, which requires limited filesystem access. (4) Verify the API domain and organization — the skill metadata has no homepage or known owner; if you need higher assurance, confirm the service's legitimacy (official docs, privacy policy) before providing videos or credentials. Autonomous invocation is allowed by default; if you want to limit automatic uploads, restrict agent permissions or avoid granting NEMO_TOKEN to the environment.

Type: OpenClaw Skill Name: screen-recording-online Version: 1.0.0 The skill facilitates cloud-based video editing via the 'nemovideo.ai' API, but it exhibits several risky behaviors that warrant caution. It automatically acquires anonymous authentication tokens, uploads user files to a remote server, and fingerprints the user's environment by probing for specific installation paths (e.g., ~/.cursor/skills/ or ~/.clawhub/) to set attribution headers. Furthermore, it implements a remote-control mechanism where the backend can dictate agent actions through SSE events (e.g., translating 'click button' into API calls), which, while plausibly needed for the service, represents a significant attack surface (SKILL.md).