MemoryMesh SDK for durable runtime memory, checkpoints, recovery, and adapters for agent frameworks.
Project description
MemoryMesh Python SDK
MemoryMesh gives AI agents durable working memory, inspectable run receipts, checkpoints, and recovery state. The Python SDK is dependency-free by default and works in agent workers, notebooks, internal automation, or backend services.
Install
pip install memorymesh-sdk
Quick Start
import os
from memorymesh import MemoryMeshClient
client = MemoryMeshClient(
base_url="https://api-two-blue-75.vercel.app",
api_key=os.environ.get("MEMORYMESH_API_KEY"),
default_memory_backend="cognee_cloud",
)
receipt = client.run_agent(
agent_id="research",
task="Compare durable memory options for coding agents.",
)
print(receipt["final_output"])
print(receipt["memory_operations"])
Memory
client.remember(
text="The build agent should preserve project constraints before editing.",
dataset="agent-lessons",
)
matches = client.recall(
query="What should the build agent remember before editing?",
dataset="agent-lessons",
top_k=3,
)
Auth
The SDK defaults to the production MemoryMesh API header, X-MemoryMesh-API-Key.
client = MemoryMeshClient(
base_url="https://your-memorymesh-api.example.com",
api_key=os.environ["MEMORYMESH_API_KEY"],
)
For gateways that expect bearer auth:
client = MemoryMeshClient(
base_url="https://your-memorymesh-api.example.com",
api_key=os.environ["MEMORYMESH_API_KEY"],
api_key_header="Authorization",
)
Tool Tracing
from memorymesh import MemoryMeshClient, ToolWrapperConfig, trace_tool
client = MemoryMeshClient("http://localhost:8000")
run = client.start_run(
agent_id="support-agent",
task="Investigate payment failures.",
)
@trace_tool(
client,
run["task_id"],
ToolWrapperConfig(
tool_name="fetch_tickets",
validation=lambda args, kwargs, result: {"records": len(result)},
checkpoint_after=True,
),
)
def fetch_tickets(status: str):
return [{"id": "ticket_1", "status": status}]
fetch_tickets("open")
Error Handling
from memorymesh import MemoryMeshError
try:
client.memory_status("cognee_cloud", probe=True)
except MemoryMeshError as error:
print(error.status, error.detail or error.body)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file memorymesh_sdk-0.1.0.tar.gz.
File metadata
- Download URL: memorymesh_sdk-0.1.0.tar.gz
- Upload date:
- Size: 9.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b3cfb5c72e817040eb388d8a1142bd6b164fb0a409f5e745c593ef512e608459
|
|
| MD5 |
b018a0870031fdd6da97956f72e180b3
|
|
| BLAKE2b-256 |
8949780195b79298118f986481e4b94697b96c346543b156098f11a26aabbcf7
|
File details
Details for the file memorymesh_sdk-0.1.0-py3-none-any.whl.
File metadata
- Download URL: memorymesh_sdk-0.1.0-py3-none-any.whl
- Upload date:
- Size: 11.5 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ab50b0438013aceb11ad97a593c1e7bc4fb15f92609a750e348e1b90442e0785
|
|
| MD5 |
9a9845d930d62cffb0c9b6ee10ebba34
|
|
| BLAKE2b-256 |
3d418754167c6b0c65d080bf8c599b6163f97fe440fb4ae5e80f764585ddb503
|