SparkAnalyzer

# SparkAnalyzer ## A Minecraft plugin that analyzes Spark profiler data using AI to provide actionable performance recommendations. ### Features **AI-Powered Analysis** - Supports multiple AI providers: OpenAI, Anthropic (Claude), Google Gemini, and OpenRouter - Evidence-based diagnostics with severity ratings (Critical, High, Medium, Low) - Structured analysis reports with specific recommendations **Easy to Use** - Simply run `/sparkanalyze <spark-url>` with any spark.lucko.me profile link - Upload raw JSON data with `/sparkanalyze upload` - Re-analyze last profile with `/sparkanalyze last` **Webhook Integration** - Send analysis reports to Discord, Slack, or custom webhooks - Automatic retry with exponential backoff - Auto-disable on repeated failures **Internationalization** - Full i18n support with customizable language files - English (en-US) included by default **Security Hardened** - No backdoors, telemetry, or hidden network calls - API keys are never logged - All network activity is user-triggered only <details> <summary>Spoil config.yml</summary> ``` # SparkAnalyzer Configuration # Security-hardened Minecraft Paper plugin for AI-powered Spark analysis # Language setting - uses standard locale codes # Available: en-US (default), vi-VN, ja-JP, etc. # If an invalid code is provided, falls back to en-US language: en-US # AI Provider Configuration ai: # Supported providers: openai, anthropic, gemini, provider: openrouter # Your API key - NEVER share this publicly # This value is loaded securely and never logged api-key: your_api_key_here # Model to use for analysis # OpenAI: gpt-4.1-mini, gpt-4o, gpt-4-turbo # Anthropic: claude-3-opus, claude-3-sonnet # Gemini: gemini-pro # OpenRouter: Any model from openrouter.ai (e.g., openai/gpt-4o, anthropic/claude-3.5-sonnet, meta-llama/llama-3.1-70b-instruct) model: liquid/lfm-2.5-1.2b-thinking:free # Temperature controls randomness (0.0 = deterministic, 2.0 = creative) # For technical analysis, lower values (0.1-0.3) are recommended # Valid range: 0.0 to 2.0 temperature: 0.2 # Maximum tokens in the AI response # Higher values allow for more detailed analysis # Must be greater than 0 max-tokens: 32000 # Webhook Configuration (Optional) webhook: # Set to true to enable webhook notifications enabled: false # Webhook URL - only this URL will receive notifications # Example: https://discord.com/api/webhooks/... url: "" # Format: discord, slack, or generic format: discord # Include full analysis report in webhook (may be truncated) include-full-report: false # Maximum retry attempts for failed webhooks max-retries: 3 # Auto-disable webhook after this many consecutive failures failure-threshold: 5 # Report Settings reports: # Save detailed reports to plugins/SparkAnalyzer/reports/ save-to-file: true # Include timestamps in report filenames include-timestamp: true # Debug Settings (for troubleshooting only) debug: # Enable verbose logging (does NOT log API keys) enabled: false ``` </details> !(https://cdn.modrinth.com/data/cached_images/d4c1bc669e79d8cc93ad2b0104d650fbddcf3422.png) !(https://cdn.modrinth.com/data/cached_images/4360ad0bce5d86519c2e5fb550d552a0d525b3d0.png) ### Requirements - Paper 1.16.5 - 1.21.11 - Java 17+ - Spark profiler installed ### Permissions | Permission | Description | |------------|-------------| | `sparkanalyzer.analyze` | Use analysis commands | | `sparkanalyzer.reload` | Reload configuration |