Python SDK

Native Python client for programmatic memory management. Use it in custom agents, scripts, or anywhere you need memory access from Python.

Installation

pip install cortex-ai-mem

Quick Start

from cortex_memory import CortexClient

client = CortexClient(api_key="ctx_your_key_here")

# Save a memory
memory = client.save(
    content="User prefers Python and dark mode",
    tags=["preference"],
    importance=0.8
)

# Search memories
results = client.search("user preferences", limit=5)
for memory in results:
    print(f"{memory.content} (score: {memory.score})")

# Update a memory
client.update(
    memory_id="mem_abc123",
    content="User prefers Python, dark mode, and VS Code"
)

# Delete a memory
client.delete(memory_id="mem_abc123")

Configuration

The client can be configured via constructor arguments or environment variables:

Parameter	Env Variable	Description
api_key	CORTEX_API_KEY	Your API key (required)
base_url	CORTEX_API_URL	API base URL (default: https://cortex-mem.com/api)

API Reference

client.save(content, tags=None, agent_id=None, importance=0.5, session_id=None)

Save a new memory. Returns the created memory object.

client.search(query, limit=10, tags=None, agent_id=None)

Search memories by semantic similarity. Returns a list of memory objects with scores.

client.update(memory_id, content=None, tags=None)

Update an existing memory's content or tags.

client.delete(memory_id)

Permanently delete a memory by ID.

Error Handling

from cortex_memory import CortexClient, CortexError

client = CortexClient(api_key="ctx_your_key_here")

try:
    results = client.search("user preferences")
except CortexError as e:
    print(f"API error: {e.status_code} - {e.message}")