Redis

Redis for state and caching

Redis

Redis is used for state management, caching, and vector search.

Installation

bash

pip install "praisonaiagents[tools]"

Docker Setup

bash

docker run -d --name praison-redis -p 6379:6379 redis:7

Quick Start

python

from praisonaiagents import Agent

agent = Agent(
    name="Assistant",
    instructions="You are a helpful assistant.",
    memory={
        "backend": "redis",
        "db": "redis://localhost:6379",
        "session_id": "my-session"
    }
)

response = agent.start("Hello!")
print(response)

Connection String Format

redis://localhost:6379
redis://:password@host:port/db_number

Environment Variables

bash

export PRAISON_STATE_URL="redis://localhost:6379"

python

import os
from praisonaiagents import Agent

agent = Agent(
    name="Assistant",
    memory={"db": os.getenv("PRAISON_STATE_URL")}
)

CLI

bash

# Validate connection
praisonai persistence doctor \
    --state-url "redis://localhost:6379"

Use Cases

Session state caching
Rate limiting
Temporary data storage
Vector search (Redis Stack)

When to Use

✅ Good for:

Session caching
Rate limiting
Real-time features
Pub/sub messaging

❌ Not recommended for:

Long-term storage
Complex queries
Large datasets

Storage Backend (Advanced)

For training data, sessions, and general persistence, use the RedisBackend:

BaseJSONStore

from praisonaiagents.storage import RedisBackend, BaseJSONStore

backend = RedisBackend(
    url="redis://localhost:6379",
    prefix="praison:",
    ttl=3600  # 1 hour TTL
)
store = BaseJSONStore("session.json", backend=backend)
store.save({"messages": ["Hello"]})
data = store.load()

TrainingStorage

from praisonai.train.agents.storage import TrainingStorage
from praisonaiagents.storage import RedisBackend

with TrainingStorage(
    session_id="train-123",
    backend=RedisBackend(url="redis://localhost:6379", prefix="train:", ttl=86400),
) as storage:
    storage.save_iteration(...)
# Redis connection closed automatically at the end of the with block

SessionManager

from praisonai.cli.state.sessions import SessionManager
from praisonaiagents.storage import RedisBackend

backend = RedisBackend(url="redis://localhost:6379", prefix="session:", ttl=3600)
manager = SessionManager(backend=backend)
sessions = manager.list(limit=10)

RunHistory

from praisonai.recipe.history import RunHistory
from praisonaiagents.storage import RedisBackend

backend = RedisBackend(url="redis://localhost:6379", prefix="runs:")
history = RunHistory(backend=backend)

MCPToolIndex

from praisonai.mcp_server.tool_index import MCPToolIndex
from praisonaiagents.storage import RedisBackend

backend = RedisBackend(url="redis://localhost:6379", prefix="mcp:")
index = MCPToolIndex(backend=backend)

CLI Usage

bash

# Training with Redis backend
praisonai train agents --input "Hello" --storage-backend redis://localhost:6379

# Session list with Redis backend
praisonai session list --storage-backend redis://localhost:6379

RedisBackend Features

Feature	Description
TTL Support	Automatic expiration with `ttl` parameter
Key Prefixing	Namespace isolation with `prefix` parameter
Sub-ms Latency	High-speed caching
Distributed	Shared state across nodes

Production notes

RedisStateStore.scan_prefix(prefix) iterates the keyspace with a non-blocking SCAN cursor (batches of 500) instead of the O(N) blocking KEYS command. Prefer it over keys(pattern) when listing many keys on a production Redis instance, so a single read never freezes other clients on the single-threaded server.

python

from praisonai.persistence import create_state_store

store = create_state_store("redis", url="redis://localhost:6379")
recent = store.scan_prefix("session:")

See Storage Backends for more details.