Subtitle Translator

Translate SRT subtitle files using LLM APIs with OpenAI-compatible format. Supports both single-language and bilingual output. Use when you need to translate...

Security Scan

The skill's code and instructions align with its stated purpose (translating SRT files via an OpenAI-compatible API); it requires an API key and network access which is expected, and there is no evidence of hidden endpoints or unrelated privileges.

Purpose & Capability

Name/description match the included scripts: parse/validate SRT files and call an OpenAI-compatible chat/completions endpoint to produce translations. No unrelated binaries, services, or capabilities are requested.

Instruction Scope

SKILL.md instructions correspond to what the scripts do: read an input .srt, validate, batch text to an LLM API, and write an output .srt. The scripts only reference SRT input/output, API URL/key/model, and standard proxy env vars; they do not attempt to read other system credentials or unrelated files.

Install Mechanism

No install spec; this is instruction + script files only. Nothing is downloaded or executed from external URLs during install, so install risk is low.

Credentials

The skill legitimately needs an API endpoint and API key to function (scripts require --api-key / env SUBTITLE_API_KEY). However, the registry metadata declared no required environment variables or primary credential — this is a documentation/metadata omission. The SKILL.md also recommends optionally storing the key in ~/.openclaw/skills/subtitle-translator/config.json (plaintext), which increases risk if the user follows it. The scripts respect http_proxy/https_proxy, meaning a misconfigured or malicious proxy could capture the API key (and SKILL.md appropriately warns about this).

Persistence & Privilege

always is false and the skill does not request system-wide persistence or modify other skills. It suggests writing an optional per-skill config file under the user's home directory (expected and scoped to the skill).

Assessment

This skill appears to do exactly what it says: translate .srt files by sending subtitle text to an OpenAI-compatible API. Before using it, consider: (1) you must supply an API URL and API key — the registry metadata omitted declaring this credential; treat the key like any secret. (2) Prefer environment variables or a secret manager over the suggested plaintext config file (~/.openclaw/skills/.../config.json). (3) The skill sends all subtitle text to whatever API URL you provide — do not use it for sensitive content unless you control the endpoint. (4) Proxies (http_proxy/https_proxy) are honored and can capture keys—ensure you trust your proxy. (5) Review the small Python scripts yourself (they're included) before running. If you want stricter guarantees, ask the author to declare the API key as a required credential in the registry metadata and to remove the plaintext-config recommendation.

Overview

Translate SRT subtitle files using LLM APIs. Supports OpenAI-compatible API format with customizable URL, API key, and model selection. Outputs single-language or bilingual subtitles with original timestamps preserved.

Features

• Format Support: SRT subtitles
• API Compatibility: OpenAI-compatible endpoints
• Output Modes: Single-language or bilingual (translation above original)
• Batch Processing: 50 sentences per batch (configurable) with 1-second intervals
• Progress Tracking: Detailed logging of task progress and execution
• Validation: Ensures sentence count consistency and timestamp preservation

Quick Start

Basic Translation

python3 scripts/translate_srt.py \
  --input video.srt \
  --output video_zh.srt \
  --source-lang en \
  --target-lang zh \
  --api-url https://api.openai.com/v1 \
  --api-key sk-... \
  --model gpt-4

Bilingual Output

python3 scripts/translate_srt.py \
  --input video.srt \
  --output video_bilingual.srt \
  --source-lang en \
  --target-lang zh \
  --bilingual \
  --api-url https://api.openai.com/v1 \
  --api-key sk-...

Validate SRT File

python3 scripts/validate_srt.py input.srt

List Available Models

python3 scripts/list_models.py \
  --api-url https://api.openai.com/v1 \
  --api-key sk-...

Configuration

Option 1: Environment Variables (Recommended)

export SUBTITLE_API_URL="https://api.openai.com/v1"
export SUBTITLE_API_KEY="sk-your-api-key"
export SUBTITLE_MODEL="gpt-4"

python3 scripts/translate_srt.py -i input.srt -o output.srt

Option 2: Command Line Arguments

python3 scripts/translate_srt.py \
  -i input.srt \
  -o output.srt \
  -u https://api.openai.com/v1 \
  -k sk-your-api-key \
  -m gpt-4

Option 3: Config File (Less Secure)

Create ~/.openclaw/skills/subtitle-translator/config.json:

{
  "api_url": "https://api.openai.com/v1",
  "api_key": "sk-your-api-key",
  "model": "gpt-4",
  "batch_size": 50,
  "batch_interval_ms": 1000,
  "output_mode": "single",
  "log_level": "info"
}

⚠️ Security Warning: Storing API keys in plaintext config files increases risk. Prefer environment variables or command line arguments.

Security Considerations

1. API Key Storage: Use environment variables or secure secret managers instead of plaintext config files
2. API Endpoint: Ensure you trust the API URL you provide
3. Proxy Environment: The scripts honor http_proxy/https_proxy environment variables. If your environment uses untrusted proxies, API keys could be captured
4. Network Traffic: All subtitle content is sent to the API endpoint. Do not use with sensitive/confidential content unless you control the API

Workflow

1. Parse SRT: Extract index, timecodes, and text
2. Validate: Optional validation of input format
3. Batch Translation: Send 50 sentences per request (configurable)
4. Validate: Ensure output count matches input
5. Reconstruct: Combine original timecodes with translations
6. Output: Generate SRT file

Resources

scripts/

• translate_srt.py - Main translation script
• list_models.py - List available models from API
• validate_srt.py - Validate SRT file format

references/

• srt_format.md - SRT file format specification