菜单
AI features and vector search with Syntra. Use when configuring AI models, running chat completions, generating images, creating embeddings, building semantic search, setting up RAG pipelines, or managing vector collections.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | syntra-ai |
| description | AI features and vector search with Syntra. Use when configuring AI models, running chat completions, generating images, creating embeddings, building semantic search, setting up RAG pipelines, or managing vector collections. |
All AI features route through OpenRouter, supporting 200+ models from OpenAI, Anthropic, Google, Meta, Mistral, and more.
OPENROUTER_API_KEY in environmentai_create_config — define a model configuration{
"input_modality": "text",
"output_modality": "text",
"provider": "openrouter",
"model_id": "anthropic/claude-sonnet-4-5-20250929",
"system_prompt": "You are a helpful assistant."
}
ai_list_configs — list all configsai_create_config — create a new config (model, provider, modality, system prompt)ai_get_config / ai_update_config / ai_delete_config — CRUD operations{
"config_id": "uuid-of-text-config",
"messages": [
{ "role": "system", "content": "You are a coding assistant." },
{ "role": "user", "content": "Explain closures in JavaScript." }
],
"temperature": 0.7,
"max_tokens": 1000
}
{
"config_id": "uuid-of-image-config",
"prompt": "A futuristic city at sunset, cyberpunk style",
"n": 1,
"size": "1024x1024"
}
{
"config_id": "uuid-of-embedding-config",
"input": "The quick brown fox jumps over the lazy dog"
}
Or batch: "input": ["text one", "text two", "text three"]
Syntra includes built-in vector storage powered by pgvector.
ai_create_collection — create a vector collectionai_upsert_documents — add documents (embeddings auto-generated)ai_vector_search — semantic search by text queryai_list_collections / ai_get_collection / ai_delete_collectionname, embedding_model, dimensions, distance_metric| Metric | Use case |
|---|---|
cosine | General-purpose similarity (default) |
l2 | Euclidean distance |
inner_product | When vectors are normalized |
ai_upsert_documents — insert or update documents with auto-embeddingai_delete_documents — remove by IDsai_vector_search — search by text query with optional metadata filterai_vector_search_by_vector — search by raw embedding vectorAuthentication and user management with Syntra. Use when setting up auth, configuring OAuth providers, managing users, handling email verification, password resets, OTP codes, or configuring password policies and session settings.
Database management with Syntra. Use when creating tables, defining schemas, querying records, inserting data, running SQL, or managing database structure. Covers CRUD operations, schema design, migrations, filtering, and PostgreSQL functions.
Edge functions and scheduled jobs with Syntra. Use when creating serverless functions, deploying code to the edge runtime, invoking functions by slug, setting up cron jobs, or managing scheduled tasks.
Initialize and configure a Syntra backend. Use when the user wants to set up Syntra, connect to their backend, verify the connection, configure MCP, or bootstrap a new project with Syntra as the backend.
System diagnostics and monitoring for Syntra. Use when checking backend health, viewing system info, inspecting audit logs, reviewing MCP usage statistics, listing modules, or troubleshooting connectivity issues.
File storage management with Syntra. Use when uploading files, creating buckets, downloading objects, generating signed URLs, managing file assets, or organizing stored content. Supports local and S3-compatible storage.