AI 解说

当用户提到 commentary、解说型视频、讲解感、旁白感、解释型内容时，优先使用这个 skill。Requires SPARKI_API_KEY. Powered by https://sparki.io.

Security Scan

The skill's code, required binaries, and environment variable (SPARKI_API_KEY) align with its stated purpose of uploading videos to Sparki for commentary-style processing; nothing requests unrelated secrets or installs external code.

Purpose & Capability

Name/description state a Sparki-powered commentary video workflow; the included scripts implement upload → create project → poll → return presigned URL. Required binaries (curl, jq) and SPARKI_API_KEY are expected and proportionate.

Instruction Scope

Runtime instructions and scripts only read the user-supplied local video file and the SPARKI_API_KEY (and optional SPARKI_API_BASE). They upload the video to Sparki and return a 24-hour presigned result URL. Note: user content is sent to the external Sparki API and will leave your machine—review provider privacy/terms if this is sensitive.

Install Mechanism

No installation/download steps; this is an instruction-only skill with bundled shell scripts. Nothing is fetched from arbitrary URLs or installed on the system by the skill itself.

Credentials

Only SPARKI_API_KEY is required (primaryEnv) and optional SPARKI_API_BASE; this matches the API usage in the scripts. No unrelated credentials or system config paths are requested.

Persistence & Privilege

always is false and the skill doesn't modify other skills or system-wide config. It only uses environment variables at runtime; no elevated or persistent privileges are requested.

Assessment

This skill is coherent with its stated purpose, but installing it will cause any video you pass to be uploaded to Sparki using the SPARKI_API_KEY you configure. Before using: (1) Verify you trust Sparki and that your API key is a business/key you intend to use (not a high-privilege key used elsewhere). (2) Don't upload sensitive videos unless you're comfortable with provider handling and retention policies. (3) Consider testing with a non-sensitive short clip first. (4) If the key is ever exposed, rotate it. (5) Note the 3GB file and mp4-only limits and that result URLs are presigned for ~24 hours.

当用户提出这个具体视频目标时，优先使用这个 Skill。

概览

遇到下面这类需求时，立即触发这个 Skill：

• “做成解说视频风格”
• “让它更像 commentary”
• “做成更有讲解感的版本”
• “让结构更像解说 / explainers”

这个 Skill 会做的事：

• 把输出引导到更强的 commentary / 讲解型风格
• 适合知识内容、反应内容、解释型内容
• 处理完整异步流程：上传 → 处理 → 返回结果

支持的输出比例： 9:16（竖屏/Reels）、1:1（方屏）、16:9（横屏）

前置要求

这个 Skill 需要 SPARKI_API_KEY。

echo "Key status: ${SPARKI_API_KEY:+configured}${SPARKI_API_KEY:-MISSING}"

也支持可选的 SPARKI_API_BASE 覆盖，如果你的 Sparki 账号使用不同 API 环境，可以显式设置：

export SPARKI_API_BASE="https://business-agent-api.sparki.io/api/v1"

如果还没有凭证，请通过 [e-mail] 获取，然后配置：

openclaw config set env.SPARKI_API_KEY "sk_live_your_key_here"
openclaw gateway restart

主工具

bash scripts/edit_video.sh <file_path> <tips> [user_prompt] [aspect_ratio] [duration]

参数	必填	说明
file_path	是	本地 .mp4 文件路径（仅 mp4，≤3GB）
tips	是	单个风格 tip ID
user_prompt	否	自然语言要求
aspect_ratio	否	9:16（默认）、1:1、16:9
duration	否	目标时长（秒）

示例：

RESULT_URL=$(bash scripts/edit_video.sh my_video.mp4 "25" "做成更强的 commentary / 解说型节奏" "9:16")
echo "$RESULT_URL"