Cortexa research tool·cs ml·embedding

Embedding Database Create

Create a per-collection datastore: <name>.db (SQLite) + <name>.faiss (FAISS). Embeds documents using the chosen provider (openai/azure/huggingface/local). Vectors are L2-normalized; FAISS index uses IndexFlatIP (cosine).

Livelast probed 2026-05-01 01:00Z· 100msCortexa

Status

The most recent probe succeeded. Safe to call from the agent today.

Last probed: 1d ago
Last success: 1d ago
Last latency: 100ms

Schema

JSON Schema the agent (or your API call) must match.

{
  "properties": {
    "action": {
      "const": "create_from_docs",
      "type": "string"
    },
    "database_name": {
      "description": "Collection/database name (produces <name>.db and <name>.faiss)",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "Optional human-readable description for the collection",
      "type": "string"
    },
    "documents": {
      "description": "List of document texts to embed and store",
      "items": {
        "type": "string"
      },
      "maxItems": 10000,
      "minItems": 1,
      "type": "array"
    },
    "metadata": {
      "default": [],
      "description": "Optional metadata for each document (must match length of documents if provided)",
      "items": {
        "type": "object"
      },
      "type": "array"
    },
    "model": {
      "description": "Embedding model/deployment id. Defaults: EMBED_MODEL, else provider-specific sensible default.",
      "type": "string"
    },
    "provider": {
      "description": "Embedding backend. Defaults: EMBED_PROVIDER, else by available creds (azure>openai>huggingface>local).",
      "enum": [
        "openai",
        "azure",
        "huggingface",
        "local"
      ],
      "type": "string"
    }
  },
  "required": [
    "database_name",
    "documents"
  ],
  "type": "object"
}

Examples (1)

Default Embedding Database Create call

public-safeinput

{
  "database_name": "",
  "documents": [],
  "metadata": []
}

Expected response keys: success

Anonymous-safe example. Rate-limited; no sign-in required.

Identifiers

Catalog ID: tu_embedding_database_create
Tool name: embedding_database_create
Added: 2026-05-01 00:58Z
Tags: tooluniverse

Schema

JSON Schema the agent (or your API call) must match.

{ "properties": { "action": { "const": "create_from_docs", "type": "string" }, "database_name": { "description": "Collection/database name (produces <name>.db and <name>.faiss)", "type": "string" }, "description": { "default": "", "description": "Optional human-readable description for the collection", "type": "string" }, "documents": { "description": "List of document texts to embed and store", "items": { "type": "string" }, "maxItems": 10000, "minItems": 1, "type": "array" }, "metadata": { "default": [], "description": "Optional metadata for each document (must match length of documents if provided)", "items": { "type": "object" }, "type": "array" }, "model": { "description": "Embedding model/deployment id. Defaults: EMBED_MODEL, else provider-specific sensible default.", "type": "string" }, "provider": { "description": "Embedding backend. Defaults: EMBED_PROVIDER, else by available creds (azure>openai>huggingface>local).", "enum": [ "openai", "azure", "huggingface", "local" ], "type": "string" } }, "required": [ "database_name", "documents" ], "type": "object" }