freemcplab / Directory / Playground / narrator-ai-cli-skill

● Community GridLtd-ProductDev ⚡ Instant

narrator-ai-cli-skill

Name: narrator-ai-cli-skill MCP Server
Author: GridLtd-ProductDev

by GridLtd-ProductDev · GridLtd-ProductDev/narrator-ai-cli-skill

Agent skill for the Narrator AI video-narration CLI — Claude drives end-to-end narration runs from the command line.

narrator-ai-cli-skill teaches Claude how to use narrator-ai-cli, the CLI client for the Narrator AI video narration API. It wraps the flags, common flows, and output handling so Claude can assemble narration prompts, kick off jobs, poll for completion, and place the output files in sensible locations — useful when you want to batch-narrate a series of videos without hand-running the CLI.

▶ Watch demo 📦 Install 💡 Use Cases

Why use it

Key features

Wraps narrator-ai-cli flags and subcommands
Handles job submission + polling patterns
Output-file placement conventions
Batch narration across multiple source videos
Fails loudly when API key or config is missing

Live Demo

What it looks like in practice

narrator-ai-cli-skill.replay ▶ ready

0/0

Install

Pick your client

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "narrator-ai-cli-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill",
        "~/.claude/skills/narrator-ai-cli-skill"
      ],
      "_inferred": true
    }
  }
}

Open Claude Desktop → Settings → Developer → Edit Config. Restart after saving.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "narrator-ai-cli-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill",
        "~/.claude/skills/narrator-ai-cli-skill"
      ],
      "_inferred": true
    }
  }
}

Cursor uses the same mcpServers schema as Claude Desktop. Project config wins over global.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "narrator-ai-cli-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill",
        "~/.claude/skills/narrator-ai-cli-skill"
      ],
      "_inferred": true
    }
  }
}

Click the MCP Servers icon in the Cline sidebar, then "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "narrator-ai-cli-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill",
        "~/.claude/skills/narrator-ai-cli-skill"
      ],
      "_inferred": true
    }
  }
}

Same shape as Claude Desktop. Restart Windsurf to pick up changes.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "narrator-ai-cli-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill",
        "~/.claude/skills/narrator-ai-cli-skill"
      ]
    }
  ]
}

Continue uses an array of server objects rather than a map.

~/.config/zed/settings.json

{
  "context_servers": {
    "narrator-ai-cli-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill",
          "~/.claude/skills/narrator-ai-cli-skill"
        ]
      }
    }
  }
}

Add to context_servers. Zed hot-reloads on save.

claude mcp add narrator-ai-cli-skill -- git clone https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill ~/.claude/skills/narrator-ai-cli-skill

One-liner. Verify with claude mcp list. Remove with claude mcp remove.

Use Cases

Real-world ways to use narrator-ai-cli-skill

How to narrate a single video from a brief

👤 Creators and editors using Narrator AI for VO ⏱ ~15 min beginner

When to use: You have a video and a draft voice-over script and want narration produced via the CLI.

Prerequisites

narrator-ai-cli installed and configured — Follow upstream install docs for the CLI
Skill cloned — git clone https://github.com/GridLtd-ProductDev/narrator-ai-cli-skill ~/.claude/skills/narrator-ai-cli-skill

Flow

Stage inputs

Given my video scene_01.mp4 and the VO script script.md, run narrator-ai-cli and save the output to ./narrations/.✓ Copied

→ CLI invocation with correct flags
Wait and verify

Poll until the job is done, then confirm the output file exists and has audio.✓ Copied

→ Output file present with non-zero duration

Outcome: Narration file ready to drop into your editor.

Pitfalls

Long jobs time out in the shell — Use the CLI's async mode + poll instead of a blocking call

Combine with: filesystem

Batch-narrate a series of scenes

👤 Teams with multi-scene projects ⏱ ~45 min intermediate

When to use: You have 10+ scenes with per-scene scripts and need them narrated consistently.

Flow

Walk the project structure

Find all scene_*.mp4 + scene_*.md pairs under project/ and queue narration for each.✓ Copied

→ Enumerated pairs + job queue
Run serially or in parallel

Run up to 3 jobs in parallel; rate-limit per the API allowance.✓ Copied

→ Controlled concurrency
Report results

Produce a table: scene, status, duration, output path.✓ Copied

→ Clear completion table

Outcome: Whole project narrated in one run with clear status.

Pitfalls

Hitting API limits mid-batch — Respect upstream rate limits; back off on 429

Combine with: filesystem

Combinations

Pair with other MCPs for X10 leverage

narrator-ai-cli-skill + filesystem

Find inputs automatically and drop outputs in structured folders

Walk project/, narrate every scene, output to project/narrations/.✓ Copied

narrator-ai-cli-skill + claude-shorts-skill

Add AI narration to Short clips generated by claude-shorts

For each rendered Short, narrate an intro line and mix it in.✓ Copied

Tools

What this MCP exposes

Tool	Inputs	When to call	Cost
cli-invocation	inputs (video+script) + flags	Every narration run	narrator-ai API usage
job-polling	job id	After submission	minimal
batch-orchestration	list of jobs	Multi-scene projects	depends on batch size

Cost & Limits

What this costs to run

API quota: depends on Narrator AI tier
Tokens per call: minimal LLM tokens; real cost is API minutes
Monetary: API billed by Narrator AI
Tip: Narrate final scripts only; iterate text in LLM first

Security

Permissions, secrets, blast radius

Credential storage: Narrator AI API key in env var; never inline

Data egress: Narrator AI endpoints

Uploaded video/audio passes through third-party service; check data handling policies

Troubleshooting

Common errors and fixes

Skill doesn't activate

Ensure SKILL.md is at the cloned dir root and Claude Code restarted.

Verify: ls ~/.claude/skills/narrator-ai-cli-skill/SKILL.md

CLI command not found

Install narrator-ai-cli and make sure it's on PATH where Claude Code runs.

Verify: which narrator-ai-cli

401 from Narrator API

Set NARRATOR_API_KEY (or the env var the CLI expects).

Verify: env | grep NARRATOR

Alternatives