Scenario-focused Sparki skill for faceless or no-camera-presence outputs while using the latest official Sparki setup, API-key, and upload workflow guidance.

This skill appears to implement a legitimate Sparki CLI for faceless video workflows and only needs your Sparki API key and local file access. However, there are a few mismatches to check before installing or using it: - Confirm the 'uv' requirement and install step: SKILL.md mentions running 'uv sync' but the package is Python-based and registry metadata claimed no install spec. Ask the publisher what 'uv' is and why it is required; avoid running unknown binaries until verified. - Be aware downloads: the CLI will download result URLs returned by the Sparki API and follows redirects — those URLs may be hosted on third-party CDNs. The skill's declared network domain is only agent-api.sparki.io, so verify whether your environment enforces network restrictions or whether the skill needs additional allowed domains. - Inspect code or run in a sandbox: source files are included, so you (or someone you trust) can review them or run the tool in an isolated environment before giving it your SPARKI_API_KEY. - Least privilege: only provide a Sparki API key scoped as narrowly as possible; do not supply high-privilege or unrelated credentials. If you cannot verify 'uv' or the publisher, treat this package with caution. The inconsistencies look more like sloppy packaging than overt malice, but they merit manual review.

Type: OpenClaw Skill Name: faceless-video Version: 1.0.12 The 'faceless-video' skill is a legitimate CLI wrapper for the Sparki AI video editing service. It handles API authentication, file uploads, project management, and result downloads via the 'agent-api.sparki.io' endpoint. The code (src/sparki_cli/) follows standard Python practices, uses well-known libraries (typer, httpx, pydantic), and its requested permissions in SKILL.md are strictly limited to its configuration and workspace directories.