Narrator AI CLI Skill
Teach your AI agent to create movie narration videos
Evidence shows: SKILL.md requires agent confirmation before acting (user_confirmation 2), and provides data flow mapping and privacy notes (data_flow_transparency 2). However, no explicit least-privilege principle is stated, API key handling only mentions scope without detailing storage and transmission (least_privilege 1, sensitive_data_handling 1). Dependency security only lists dependencies without version pinning or vulnerability scanning (dependency_security 1). External effects: the skill calls external APIs to create videos, but side effects or resource consumption are not described (external_effects 1). Rollback: no undo or recovery mechanism is provided (rollback 1). Source attribution: no author or maintainer info, only a contact email (source_attribution 1).
Evidence shows: README and SKILL.md are consistent, workflow descriptions are clear (self_consistency 2). Dependency availability: dependencies are listed, but no version pinning or mirrors (dependency_availability 1). Failure messages: SKILL.md mentions error handling, but no specific error message examples (failure_messages 2).
Evidence shows: README lists multiple platforms and scenarios (audience_and_scenarios 2), capability boundaries are described in SKILL.md (capability_boundaries 2), trigger words and examples are clear (trigger_precision 2), environment fit is good (environment_fit 2).
Evidence shows: information architecture is clear with table of contents and index (information_architecture 2), installation instructions are detailed (install_notes 2), naming is stable (naming_stability 1), examples and FAQ exist (examples_and_faq 2), known limitations are partially mentioned (known_limitations 1), MIT license is explicit (license 2), versioning and changelog are missing (versioning_changelog 1), maintenance responsibility is not clear (maintenance_responsibility 1).
Evidence shows: output usability is high, provides download links (output_usability 2), marginal value is high, automated video production (marginal_value 2), cost-benefit: requires API key, but no pricing info (cost_benefit 1).
Evidence shows: claims have partial support, but no tests or verification (claim_traceability 1), cross-source verification is insufficient (cross_source_corroboration 1), fact-inference separation is not clear (fact_inference_separation 1).
- API key handling details are insufficient; recommend clarifying storage and transmission.
- Dependencies are not version-pinned, posing supply chain risk.
- No rollback mechanism provided; recovery after failure is difficult.
- No versioning or changelog, poor maintainability.
What does this agent do, and when should you use it?
Narrator AI CLI Skill is a machine-readable skill file (SKILL.md) that teaches AI agents (such as OpenClaw, Windsurf, Claude Code, etc.) how to use the narrator-ai-cli command-line tool to automate the production of movie narration videos—from script to final video. It bundles usage instructions, resource lookup tables, workflow paths, error handling, and cost estimation, allowing agents to drive video creation through natural language commands. It depends on narrator-ai-cli v1.0.0+ and Python 3.10+, and requires an API key to access the backend service. It has been verified on 8 platforms including OpenClaw, Windsurf, and WorkBuddy. The skill package itself does not contain a video generation engine; it acts as a 'recipe book' teaching the AI how to use the 'kitchen tools' (CLI).
Reads SKILL.md and the references/ directory containing resource tables, workflows, and operational rules, and executes a series of narrator-ai-cli commands to produce narration videos: selects a movie from the catalog, picks BGM, voice, and template, generates a script, composes the video, and returns a download link. It supports two workflow paths (Adapted Narration and Original Narration) and three creation modes (Hot Drama / Original Mix / New Drama), and can perform standalone tasks like voice cloning and text-to-speech. The agent must follow Agent Rules: confirm before acting, language chain, polling pattern, etc. All operations interact with a remote API via the CLI, producing video files and file IDs.
- Content creators who want to batch-produce narration videos for popular dramas by simply saying 'Create a narration video for The Shawshank Redemption'—the agent auto-selects movies, BGM, voice, and composes the video.
- Short-video operators needing differentiated narration styles for different action movies, specifying a comedy template via natural language to generate multiple videos.
- Individual developers configuring this skill in OpenClaw to trigger video creation via voice commands, streamlining manual operations.
- Educators using Original Narration mode to create explanatory videos for historical or literary works, leveraging 90+ templates and 63 voices.
- Users needing voice cloning or text-to-speech, where the agent can directly invoke narrator-ai-cli standalone tasks for custom narration.
What are this agent's strengths and limitations?
- Provides a complete end-to-end video creation pipeline, from script to final video, fully automated by the agent.
- Bundles rich resources: ~100 movies, 146 BGM tracks, 63 voices, 90+ narration templates.
- Covers 8 major agent platforms including OpenClaw, Windsurf, Claude Code, Cursor, offering broad compatibility.
- Includes detailed error handling with all 18 API error codes and recommended actions for troubleshooting.
- Tightly couples to narrator-ai-cli and backend API, requiring API key acquisition and Python 3.10+.
- The skill is only instructions and documentation; it cannot work standalone and must be used with the CLI.
- API key procurement requires contacting a specific email, potentially involving approval or cost.
- Real-world results depend on the backend service quality and stability, possibly leading to vendor lock-in.
How do you install or deploy this agent?
- Install CLI: pip install "narrator-ai-cli @ git+https://github.com/NarratorAI-Studio/narrator-ai-cli.git"
- Configure API key: narrator-ai-cli config set app_key <your_app_key> (obtain from [email protected])
- Clone the skill repo into your agent's skills directory. For OpenClaw: mkdir -p ~/.openclaw/skills && git clone https://github.com/NarratorAI-Studio/narrator-ai-cli-skill.git ~/.openclaw/skills/narrator-ai-cli
For Windsurf/Claude Code: clone into project's .skills; for Cursor: .cursor/rules; for any markdown-skill agent, use any path.
For WorkBuddy/QClaw: upload SKILL.md and the references/ folder via UI, keeping structure intact.
How do you use this agent?
After installation, issue natural language commands in the supported agent platform, e.g.:
"Create a narration video for The Shawshank Redemption".
The agent follows the SKILL.md workflow: search movie, select template, choose BGM, pick voice, generate script, compose video, and return download link. You can also query available movies ("Show me what movies are available") or batch create ("Make 5 narration videos for different action movies"). Ensure the CLI is installed and API key is configured.