Translation (Azure AI Translator)

Reviewed 2026-10-13codingPrompt / Rule# translation # azure # translator # python # localization

Microsoft's official skill that teaches coding agents to translate text with the Azure AI Translator Python SDK: translate, detect language, transliterate and look up dictionary entries.

Security Reviewed

Low risk: a Markdown-only SDK guide; any network traffic goes to your own Azure Translator resource from code you ask the agent to write.

Install Translation (Azure AI Translator) in Google Antigravity

Antigravity loads skills from .agents/skills/ in your project or ~/.gemini/config/skills/ for all projects (official docs). Full guide

EasiestOption 1: paste this to your Antigravity agent

Install the "Translation (Azure AI Translator)" skill for Google Antigravity.
Follow the steps at https://www.antigravityskills.org/install/translation.md (source: https://github.com/microsoft/skills/tree/main/.github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py).
Before running anything, show me what you will download and whether it includes scripts, and wait for my OK.
Install it into this project (.agents/skills/) unless I say global. When done, check that Antigravity lists it as an installed skill.

The agent reads the install file, shows you what it will download and waits for your OK.

Option 2: install it yourself

1. Clone the source repository
```
git clone --depth 1 https://github.com/microsoft/skills /tmp/ms-skills
```
Clone to a temporary folder so you can copy the skill out of it.

2. Copy into this project

mkdir -p .agents/skills && cp -r /tmp/ms-skills/.github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py .agents/skills/

Run from your project root. Antigravity reads workspace skills from .agents/skills/<folder>/.

3. Or copy it for all projects

mkdir -p ~/.gemini/config/skills && cp -r /tmp/ms-skills/.github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py ~/.gemini/config/skills/

Global skills live in ~/.gemini/config/skills/.

4. Check that Antigravity sees it
Start a new session and ask "What skills do you have?" (Antigravity CLI: /skills). azure-ai-translation-text-py should appear.

The microsoft/skills README does not mention Antigravity; this is a manual copy of the skill folder. The code it produces needs pip install azure-ai-translation-text (plus azure-identity for Entra ID) and an Azure Translator resource with its endpoint, or key and region, in environment variables.

Using Claude Code, Cursor or Codex? See other assistants below.

What is Translation (Azure AI Translator)?

azure-ai-translation-text-py is a skill from Microsoft's microsoft/skills repository (SKILL.md metadata author: Microsoft, version 1.0.0, MIT) for Python developers who want an agent to add text translation to an app or script through Azure AI Translator. It does not translate by itself: it gives the agent correct, current patterns for the azure-ai-translation-text package, so the code it writes calls your own Azure Translator resource. We counted the folder at the reviewed commit: 3 Markdown files and about 17.8 KB (SKILL.md 10,880 bytes plus references/capabilities.md and references/non-hero-scenarios.md), 13 Python snippets in SKILL.md, a table of 7 client methods and 9 best practices. It is one of 40 skills in the repository's azure-sdk-python plugin; sibling skills cover document translation (azure-ai-translation-document-py) and TypeScript (azure-ai-translation-ts).

How to use Translation (Azure AI Translator)?

Install the skill, then ask the agent for translation code; the frontmatter triggers on phrases such as "text translation", "translate text", "transliterate" and "TextTranslationClient". The agent will pip install azure-ai-translation-text, build a TextTranslationClient inside a with block, and read the endpoint and credentials from environment variables. By default it uses DefaultAzureCredential (Entra ID), which needs a custom-subdomain endpoint (https://<resource>.cognitiveservices.azure.com) in AZURE_TRANSLATOR_ENDPOINT; for keyed deployments on the global endpoint it uses AZURE_TRANSLATOR_KEY plus AZURE_TRANSLATOR_REGION. Ask for batching (up to 100 texts per request, per the skill's best practices), a known source language, HTML text_type or the async client when you need them.

Key features of Translation (Azure AI Translator)

✓Translate one or many texts into one or several target languages in one call
✓Language detection with a confidence score (shown through translate results)
✓Transliteration between scripts, for example Latin to Japanese
✓Dictionary lookup and dictionary usage examples
✓Options: HTML text type, profanity handling and markers, word alignment, sentence lengths
✓Sync and async clients, Entra ID (DefaultAzureCredential) and legacy key-plus-region auth

When to use Translation (Azure AI Translator)

1Prompt from the repository's own test scenarios (tests/scenarios/azure-ai-translation-text-py): "Create a basic example that translates English text to Spanish and French. Include proper authentication with Entra ID credentials, client initialization, and proper handling of the translation results." The expected code uses DefaultAzureCredential, TextTranslationClient and translate(..., to=["es", "fr"]).
2Test-scenario prompt: "Create an example that converts text from one script to another using transliteration. Show converting Japanese text written in Latin characters to Japanese script." Expect client.transliterate with from_script "Latn" and to_script "Jpan".
3Our example: "Write a script that reads strings.json and translates every value into es, fr, de and ja with Azure AI Translator, batching up to 100 strings per request." Expect a with-block client, batched translate calls and output files per language; your Azure resource is billed for the characters sent.

Installation & Setup Guide

Agent Rules / Prompt Setup

Follow these steps (taken from the project's own docs) to configure Translation (Azure AI Translator) in your development environment.

Step 1: skills CLI (any supported agent)Claude Code / Cursor / Codex

npx skills add microsoft/skills

README, Quick Start: pick azure-ai-translation-text-py in the wizard and choose your agent; the skill is copied to that agent's folder (the README's example is .github/skills/ for GitHub Copilot). The README does not name Claude Code, Cursor or Codex individually.

Step 2: Manual copy (git clone)Any

git clone https://github.com/microsoft/skills.git && cp -r skills/.github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py <your-agent-skills-folder>/

README, Quick Start, Manual installation: clone and copy the skill folders you need. .github/skills/azure-ai-translation-text-py in the repo is a symlink to this plugin folder.

Step 3: Claude Code via symlinkClaude Code

ln -s ../.github/skills .claude/skills

README, Manual installation: share a project's .github/skills folder with Claude Code by symlinking it as .claude/skills. Cursor and Codex paths are not documented in the README.

Usage Example & Verification

Prompt / Request Example

Create a basic example that translates English text to Spanish and French. Include proper authentication with Entra ID credentials, client initialization, and proper handling of the translation results. (from the repository's test scenarios)

Expected Output / Response

Python that installs azure-ai-translation-text and azure-identity, creates DefaultAzureCredential(require_envvar=True), opens TextTranslationClient(endpoint=os.environ["AZURE_TRANSLATOR_ENDPOINT"], credential=...) in a with block, calls client.translate(body=[...], to=["es", "fr"]) and prints each translation.text and translation.to.

Installation Confirmation Method

Ask "What skills do you have?" in Antigravity (Antigravity CLI: /skills) (or ask which skills are loaded): azure-ai-translation-text-py should be listed. Then ask "translate text to German with Azure"; the agent should mention TextTranslationClient and AZURE_TRANSLATOR_ENDPOINT rather than a generic HTTP call.

Security Review

Static review

Reviewed Date

2026-10-13

Audited Ref / Commit

18eb319b121b92ea19ca911fc3bd4cc53769bb1c (2026-10-05)

Reviewer

antigravityskills.org (static code review)

Low risk: a Markdown-only SDK guide; any network traffic goes to your own Azure Translator resource from code you ask the agent to write.

Full review: scope, findings, risks

Audit Scope:.github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py/ at commit 18eb319 (SKILL.md, references/capabilities.md, references/non-hero-scenarios.md; unchanged up to HEAD 91ec095 on 2026-10-12); the .github/skills symlink; tests/scenarios/azure-ai-translation-text-py/scenarios.yaml; repository README.md Quick Start and LICENSE.

Key Findings:Documentation only, no executable files. It tells the agent to run pip install azure-ai-translation-text and to read AZURE_TRANSLATOR_ENDPOINT, AZURE_TOKEN_CREDENTIALS, AZURE_TRANSLATOR_KEY and AZURE_TRANSLATOR_REGION from the environment. The only service it targets is Azure Translator (your custom endpoint or api.cognitive.microsofttranslator.com).

Identified Operational Risks & Permissions:Text you translate is sent to Azure and billed there. If you use the legacy key path, keep the key in environment variables, as the skill shows. The parameter-name mismatch between SKILL.md and the reference file can produce code that fails on your SDK version. Static review only; nothing was run.

Static review on 2026-10-13 of microsoft/skills at commit 18eb319b121b92ea19ca911fc3bd4cc53769bb1c, skill folder .github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py (3 Markdown files, 17,798 bytes, MIT). SKILL.md instructs installing azure-ai-translation-text with pip and building TextTranslationClient inside context managers, with DefaultAzureCredential(require_envvar=True) and AZURE_TOKEN_CREDENTIALS=prod in production, or AzureKeyCredential with AZURE_TRANSLATOR_KEY and AZURE_TRANSLATOR_REGION for keyed deployments. There are no scripts, no hooks, no instructions to write files beyond the code you request and no endpoints other than Azure Translator. Static inspection only: we did not run the generated code.

Operational Notes & Known Limitations

Known Limitations

•Needs an Azure Translator resource; usage is billed by Azure. Entra ID auth needs a custom-subdomain endpoint; the global endpoint needs a key and region.
•Python only. Use the sibling azure-ai-translation-ts (TypeScript) or azure-ai-translation-document-py (whole documents) skills for those cases.
•The two files disagree (we compared them): SKILL.md calls lookup_dictionary_entries with from_parameter= and to= and plain strings, while references/non-hero-scenarios.md uses from_language=, to_language= and InputTextItem objects. Check the parameter names against the SDK version you install.
•The client-method table lists detect, but no snippet calls it; language detection is shown only through translate results.
•Microsoft tests the skill (we counted 11 scenarios in tests/scenarios/azure-ai-translation-text-py, run with gpt-5.5 in the test config); passing those checks means the code matches expected patterns, not that it ran against Azure.

Frequently Asked Questions

Is the Azure AI Translation skill safe?

In our static review the folder holds only Markdown, no scripts. Code the agent writes from it sends your text to your Azure Translator resource and reads credentials from environment variables; the skill recommends Entra ID over API keys.

How do I install it in Antigravity?

Clone microsoft/skills and copy .github/plugins/azure-sdk-python/skills/azure-ai-translation-text-py into .agents/skills/ or ~/.gemini/config/skills/, then ask "What skills do you have?" (Antigravity CLI: /skills).

What do I need on Azure?

A Translator resource. For Entra ID auth, its custom-subdomain endpoint in AZURE_TRANSLATOR_ENDPOINT; for key auth on the global endpoint, AZURE_TRANSLATOR_KEY and AZURE_TRANSLATOR_REGION.

Can it translate whole documents?

No, this skill covers text. The same repository has azure-ai-translation-document-py for documents.

Supported Platforms

antigravityclaudecodex

View Source on GitHub

Keep exploring this topic

Browse the full Antigravity skills directory for more source-linked skills with install steps and static safety reviews.

More coding skills #translation #azure #translator #python Use cases

Related Skills

✍️

Technical Writing

@anthropics

Anthropic's doc-coauthoring skill: a three-stage workflow in which the agent interviews you, drafts a spec or design doc section by section, then tests it on a fresh reader.

Static reviewwrites filesinstructions only

technical-writingdocumentation

🎨