hindsight-google-adk

Persistent long-term memory for Google ADK agents via Hindsight.

The package gives you two complementary patterns:

• HindsightMemoryService — Implements ADK's BaseMemoryService. Pass it to Runner(memory_service=...) and sessions are automatically retained on completion; agents calling search_memory get results back from Hindsight.
• create_hindsight_tools(...) — Returns a list of ADK FunctionTool instances (hindsight_retain, hindsight_recall, hindsight_reflect) the model can call inside a turn.

Installation

pip install hindsight-google-adk

Quick Start: Automatic Memory

import asyncio
from google.adk.agents import LlmAgent
from google.adk.runners import Runner
from google.adk.sessions import InMemorySessionService

from hindsight_google_adk import HindsightMemoryService

memory = HindsightMemoryService.from_url(
    hindsight_api_url="https://api.hindsight.vectorize.io",
    api_key="hsk_...",
)

agent = LlmAgent(name="assistant", model="gemini-2.0-flash")

runner = Runner(
    app_name="my-app",
    agent=agent,
    session_service=InMemorySessionService(),
    memory_service=memory,
)

When a session ends, Runner calls add_session_to_memory, which retains the session's events to a Hindsight bank keyed by (app_name, user_id). When the agent (or another call) invokes search_memory(app_name, user_id, query), the integration runs a Hindsight recall and returns the results as ADK MemoryEntry objects.

Quick Start: Explicit Tools

from google.adk.agents import LlmAgent
from hindsight_google_adk import create_hindsight_tools

tools = create_hindsight_tools(
    bank_id="user-123",
    hindsight_api_url="https://api.hindsight.vectorize.io",
    api_key="hsk_...",
)

agent = LlmAgent(name="assistant", model="gemini-2.0-flash", tools=tools)

The agent now has three tools (toggle with include_retain / include_recall / include_reflect):

• hindsight_retain(content) — store information to long-term memory
• hindsight_recall(query) — search memory and return matches
• hindsight_reflect(query) — synthesize a coherent answer from memory

Global Configuration

For shared defaults, call configure(...) once at startup:

from hindsight_google_adk import configure

configure(
    hindsight_api_url="https://api.hindsight.vectorize.io",
    api_key=None,                      # falls back to HINDSIGHT_API_KEY env var
    bank_id_template="{app_name}::{user_id}",
    budget="mid",
    max_tokens=4096,
)

Subsequent HindsightMemoryService.from_url() and create_hindsight_tools() calls use the global config as a fallback.

Bank ID Derivation

By default, each (app_name, user_id) pair gets its own bank: "{app_name}::{user_id}". Override with bank_id_template:

# Per-user bank shared across apps
HindsightMemoryService.from_url(..., bank_id_template="user::{user_id}")

# Static bank shared across all users
HindsightMemoryService.from_url(..., bank_id_template="my-shared-bank")

Configuration Reference

Memory Scoping with Tags

Each retained session/event carries app:<app_name> and user:<user_id> tags. Recall queries automatically include user:<user_id> so users never see each other's memories. Pass tags=[...] to extend the retain set or recall_tags=[...] to filter recall further.

Connection Modes

Hindsight Cloud (recommended)

HindsightMemoryService.from_url(
    hindsight_api_url="https://api.hindsight.vectorize.io",
    api_key="hsk_...",
)

Self-hosted

HindsightMemoryService.from_url(hindsight_api_url="http://localhost:8888")

Error Handling

All add_* and search_memory methods are resilient: Hindsight failures are logged but never propagate to the Runner. The explicit tools raise HindsightError on failure so the agent can react.

Requirements

• Python 3.10+
• google-adk>=2.0
• hindsight-client>=0.4.0

License

MIT