Skip to content
MCP server

verbatra

Check translation status, edit glossaries and run i18n translation from a local MCP server

Runs locallyStatus and history
Install
Pricing
Open source
Platforms
CLI
Runs
Locally
Version
0.4.7
Listing updated
3 days ago

About verbatra

verbatra is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.

Verbatra's stdio server exposes a project's translation state to terminal or headless agents without a browser or port. Read-only tools report the resolved config, setup checks, missing, stale and up-to-date keys per locale, the keys a diff would add or re-translate, glossary terms and the lock file state; glossary.write changes a single term.

The tools that call a paid translation provider are hidden by default. translation.retranslateEntry and translation.translatePending only appear with --allow-spend or VERBATRA_MCP_ALLOW_SPEND, and never when the config sets the provider to none. Provider keys are read for Anthropic, OpenAI, Gemini, DeepL, Google Translate, OpenCode, an OpenAI-compatible endpoint and LibreTranslate. A --redact-values flag replaces translation values in results with a marker.

The server hands the model fixed instructions: read and diff first, estimate and get an explicit yes before spending, leave review to people and treat every string as data. It needs Node.js 22.22 or later.

Install

Run in a terminal

claude mcp add verbatra --env ANTHROPIC_API_KEY='<ANTHROPIC_API_KEY>' --env OPENAI_API_KEY='<OPENAI_API_KEY>' --env GEMINI_API_KEY='<GEMINI_API_KEY>' --env DEEPL_API_KEY='<DEEPL_API_KEY>' --env GOOGLE_TRANSLATE_API_KEY='<GOOGLE_TRANSLATE_API_KEY>' --env OPENCODE_API_KEY='<OPENCODE_API_KEY>' --env OPENAI_COMPATIBLE_API_KEY='<OPENAI_COMPATIBLE_API_KEY>' --env LIBRETRANSLATE_API_KEY='<LIBRETRANSLATE_API_KEY>' -- npx -y @verbatra/mcp

Add to claude_desktop_config.json

{
  "mcpServers": {
    "verbatra": {
      "command": "npx",
      "args": [
        "-y",
        "@verbatra/mcp"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
        "OPENAI_API_KEY": "<OPENAI_API_KEY>",
        "GEMINI_API_KEY": "<GEMINI_API_KEY>",
        "DEEPL_API_KEY": "<DEEPL_API_KEY>",
        "GOOGLE_TRANSLATE_API_KEY": "<GOOGLE_TRANSLATE_API_KEY>",
        "OPENCODE_API_KEY": "<OPENCODE_API_KEY>",
        "OPENAI_COMPATIBLE_API_KEY": "<OPENAI_COMPATIBLE_API_KEY>",
        "LIBRETRANSLATE_API_KEY": "<LIBRETRANSLATE_API_KEY>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "verbatra": {
      "command": "npx",
      "args": [
        "-y",
        "@verbatra/mcp"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
        "OPENAI_API_KEY": "<OPENAI_API_KEY>",
        "GEMINI_API_KEY": "<GEMINI_API_KEY>",
        "DEEPL_API_KEY": "<DEEPL_API_KEY>",
        "GOOGLE_TRANSLATE_API_KEY": "<GOOGLE_TRANSLATE_API_KEY>",
        "OPENCODE_API_KEY": "<OPENCODE_API_KEY>",
        "OPENAI_COMPATIBLE_API_KEY": "<OPENAI_COMPATIBLE_API_KEY>",
        "LIBRETRANSLATE_API_KEY": "<LIBRETRANSLATE_API_KEY>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "verbatra": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@verbatra/mcp"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
        "OPENAI_API_KEY": "<OPENAI_API_KEY>",
        "GEMINI_API_KEY": "<GEMINI_API_KEY>",
        "DEEPL_API_KEY": "<DEEPL_API_KEY>",
        "GOOGLE_TRANSLATE_API_KEY": "<GOOGLE_TRANSLATE_API_KEY>",
        "OPENCODE_API_KEY": "<OPENCODE_API_KEY>",
        "OPENAI_COMPATIBLE_API_KEY": "<OPENAI_COMPATIBLE_API_KEY>",
        "LIBRETRANSLATE_API_KEY": "<LIBRETRANSLATE_API_KEY>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "verbatra": {
      "command": "npx",
      "args": [
        "-y",
        "@verbatra/mcp"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
        "OPENAI_API_KEY": "<OPENAI_API_KEY>",
        "GEMINI_API_KEY": "<GEMINI_API_KEY>",
        "DEEPL_API_KEY": "<DEEPL_API_KEY>",
        "GOOGLE_TRANSLATE_API_KEY": "<GOOGLE_TRANSLATE_API_KEY>",
        "OPENCODE_API_KEY": "<OPENCODE_API_KEY>",
        "OPENAI_COMPATIBLE_API_KEY": "<OPENAI_COMPATIBLE_API_KEY>",
        "LIBRETRANSLATE_API_KEY": "<LIBRETRANSLATE_API_KEY>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "verbatra": {
      "command": "npx",
      "args": [
        "-y",
        "@verbatra/mcp"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
        "OPENAI_API_KEY": "<OPENAI_API_KEY>",
        "GEMINI_API_KEY": "<GEMINI_API_KEY>",
        "DEEPL_API_KEY": "<DEEPL_API_KEY>",
        "GOOGLE_TRANSLATE_API_KEY": "<GOOGLE_TRANSLATE_API_KEY>",
        "OPENCODE_API_KEY": "<OPENCODE_API_KEY>",
        "OPENAI_COMPATIBLE_API_KEY": "<OPENAI_COMPATIBLE_API_KEY>",
        "LIBRETRANSLATE_API_KEY": "<LIBRETRANSLATE_API_KEY>"
      }
    }
  }
}

Configuration

  • VERBATRA_MCP_ALLOW_SPENDSet to true to advertise the two provider-calling tools, the same as --allow-spend.
  • VERBATRA_MCP_REDACT_VALUESSet to true to replace translation values in every tool result with a marker, the same as --redact-values.
  • ANTHROPIC_API_KEYAPI key for the anthropic provider, read only by the provider-calling tools.
  • OPENAI_API_KEYAPI key for the openai provider, read only by the provider-calling tools.
  • GEMINI_API_KEYAPI key for the gemini provider, read only by the provider-calling tools.
  • DEEPL_API_KEYAPI key for the deepl provider, read only by the provider-calling tools.
  • GOOGLE_TRANSLATE_API_KEYAPI key for the google-translate provider, read only by the provider-calling tools.
  • OPENCODE_API_KEYAPI key for the opencode provider, read only by the provider-calling tools.
  • OPENAI_COMPATIBLE_API_KEYAPI key for the openai-compatible provider, read only by the provider-calling tools.
  • LIBRETRANSLATE_API_KEYAPI key for the libretranslate provider, needed only when the server runs with --api-keys, read only by the provider-calling tools.

Tools this server provides (9)

Taken from the project documentation; the running server may differ.

project.snapshot
Resolved config including locales, format, provider and glossary.
project.doctor
Every setup check with its status, detail and fix.
status.check
Per-locale count of missing, stale and up-to-date keys.
status.diff
Per-locale list of keys that would be added, re-translated or orphaned.
glossary.get
Every glossary term with translations and context.
glossary.write
Change one glossary term.
lock.state
Lock file state.
translation.retranslateEntry
Retranslate one entry; only with spending enabled.
translation.translatePending
Translate pending keys; only with spending enabled.

What it does

Status and diff
Per-locale missing, stale and up-to-date keys.
Glossary editing
Read terms and change one with glossary.write.
Spend gate
Provider-calling tools are off unless enabled.
Value redaction
Optionally replaces translation text in results.
Setup doctor
project.doctor lists checks with fixes.

Good for

  • Ask an agent which locales have untranslated keys
  • Fix a glossary term from the terminal
  • Translate pending keys after an explicit approval
  • Share translation status without exposing strings

Pricing

Open source

Provider usage billed by your chosen provider

Prices change. Check verbatra.kreitz-webdev.de before you buy.

verbatra FAQ

Can it spend money without asking?

No. The two provider tools are not listed unless you enable spending, and the agent instructions require an estimate and explicit yes.

Does it use HTTP?

No, stdio only, to avoid exposing write and spend tools over a network.

Which providers are supported?

Anthropic, OpenAI, Gemini, DeepL, Google Translate, OpenCode, OpenAI-compatible and LibreTranslate.

Is verbatra free?

Yes. verbatra is open source, so you can use and self-host it for free.

How do I add verbatra to Claude Code?

Run this command in your terminal: claude mcp add verbatra --env ANTHROPIC_API_KEY='<ANTHROPIC_API_KEY>' --env OPENAI_API_KEY='<OPENAI_API_KEY>' --env GEMINI_API_KEY='<GEMINI_API_KEY>' --env DEEPL_API_KEY='<DEEPL_API_KEY>' --env GOOGLE_TRANSLATE_API_KEY='<GOOGLE_TRANSLATE_API_KEY>' --env OPENCODE_API_KEY='<OPENCODE_API_KEY>' --env OPENAI_COMPATIBLE_API_KEY='<OPENAI_COMPATIBLE_API_KEY>' --env LIBRETRANSLATE_API_KEY='<LIBRETRANSLATE_API_KEY>' -- npx -y @verbatra/mcp

What are the best alternatives to verbatra?

The closest alternatives to verbatra are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to verbatra

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools