agentuity-cli-cloud-vector-upsert

Add or update vectors in the vector storage. Requires authentication. Use for Agentuity cloud platform operations

⭐3stars🔀5forks📁View on GitHub🕐Updated Jan 5, 2026⬇️1downloads

When & Why to Use This Skill

The agentuity-cli-cloud-vector-upsert skill is a powerful tool designed for managing high-dimensional vector data within the Agentuity cloud platform. It enables users to add or update vectors, document embeddings, and associated metadata, providing the essential infrastructure for Retrieval-Augmented Generation (RAG) and persistent long-term memory for AI agents.

Use Cases

  • Knowledge Base Synchronization: Automatically upserting new technical documentation or company wikis into vector storage to ensure AI agents have access to the latest information.
  • Persistent Agent Memory: Storing and updating user interaction summaries as vectors to allow agents to maintain context and recall across multiple sessions.
  • Bulk Data Indexing: Using the CLI to batch-upload large datasets (like product catalogs or research papers) from JSON files for semantic search optimization.
  • Custom Embedding Integration: Pushing pre-computed embeddings from specialized external models into the Agentuity ecosystem to support niche domain-specific search requirements.
nameagentuity-cli-cloud-vector-upsert
descriptionAdd or update vectors in the vector storage. Requires authentication. Use for Agentuity cloud platform operations
version"0.0.104"
licenseApache-2.0
allowed-tools"Bash(agentuity:*)"
argument-hint"<namespace> [key]"
command"agentuity cloud vector upsert"
tags"mutating updates-resource slow requires-auth"

Cloud Vector Upsert

Add or update vectors in the vector storage

Prerequisites

  • Authenticated with agentuity auth login
  • Project context required (run from project directory or use --project-id)

Usage

agentuity cloud vector upsert <namespace> [key] [options]

Arguments

Argument Type Required Description
<namespace> string Yes -
<key> string No -

Options

Option Type Required Default Description
--document string Yes - document text to embed
--embeddings string Yes - pre-computed embeddings as JSON array
--metadata string Yes - metadata as JSON object
--file string Yes - path to JSON file containing vectors, or "-" for stdin

Examples

Upsert a single vector with document text:

bunx @agentuity/cli vector upsert products doc1 --document "Comfortable office chair"

Upsert with metadata:

bunx @agentuity/cli vector upsert products doc1 --document "Chair" --metadata '{"category":"furniture"}'

Upsert with pre-computed embeddings:

bunx @agentuity/cli vector upsert embeddings vec1 --embeddings "[0.1, 0.2, 0.3]"

Bulk upsert from JSON file:

bunx @agentuity/cli vector upsert products --file vectors.json

Bulk upsert from stdin:

cat vectors.json | bunx @agentuity/cli vector upsert products -

Output

Returns JSON object:

{
  "success": "boolean",
  "namespace": "string",
  "count": "number",
  "results": "array",
  "durationMs": "number"
}
Field Type Description
success boolean Whether the operation succeeded
namespace string Namespace name
count number Number of vectors upserted
results array Upsert results with key-to-id mappings
durationMs number Operation duration in milliseconds