spectron-google-adk

Give Google ADK agents persistent memory backed by Spectron, SurrealDB's agent-memory layer.

An agent can store facts, search them back, forget them, and reason over them, with memory that survives restarts and separate conversations. Spectron handles entity extraction, knowledge-graph storage, temporal facts, and hybrid retrieval; this package is the thin layer that hands those verbs to an ADK agent.

Install

pip install spectron-google-adk

This pulls in google-adk and surrealdb. The Spectron client ships in the surrealdb package (3.0.0a1 and later), which is installed for you from PyPI.

Configuration

Spectron needs a context id, an endpoint, and an API key. The Spectron SDK does not read environment variables itself, so you pass these in explicitly. For scripts, SpectronConfig.from_env() reads them from the environment for you:

export SPECTRON_CONTEXT="acme-prod"
export SPECTRON_ENDPOINT="https://api.spectron.example"
export SPECTRON_API_KEY="sk-spec-..."
export GOOGLE_API_KEY="your-google-api-key"   # used by the ADK model

See .env.example for the full list.

Quickstart

import asyncio
from google.adk.agents import Agent
from google.adk.runners import InMemoryRunner
from spectron_google_adk import SpectronToolset

async def main():
    toolset = SpectronToolset(
        context="acme-prod",
        endpoint="https://api.spectron.example",
        api_key="sk-spec-...",
    )

    agent = Agent(
        model="gemini-2.5-flash",
        name="assistant",
        description="An assistant with persistent memory.",
        instruction="Store durable facts with remember and look things up with recall.",
        tools=[toolset],
    )

    runner = InMemoryRunner(agent=agent)
    try:
        await runner.run_debug("Remember: Acme Corp, healthcare, 1.2M dollar contract.")
        events = await runner.run_debug("What healthcare contracts do we have?")
        for event in events:
            if event.is_final_response() and event.content:
                for part in event.content.parts:
                    if part.text:
                        print(part.text)
    finally:
        await runner.close()
        await toolset.close()

asyncio.run(main())

SpectronToolset extends ADK's BaseToolset, so an ADK Runner closes it on shutdown. When the toolset creates the client, close() closes that client too.

Two ways to build tools

SpectronToolset (recommended)

Owns the client and manages its lifecycle. Add it to an agent as a single item in the tools list:

toolset = SpectronToolset(config=SpectronConfig.from_env())
agent = Agent(model="gemini-2.5-flash", name="assistant", tools=[toolset])

get_spectron_tools (for quick scripts)

Returns a plain list of tools. Pass your own client when you want to control its lifecycle; otherwise the client it creates lives for the process:

from surrealdb.spectron import AsyncSpectron
from spectron_google_adk import get_spectron_tools

client = AsyncSpectron(context="acme-prod", endpoint="...", api_key="sk-...")
tools = get_spectron_tools(client=client)
agent = Agent(model="gemini-2.5-flash", name="assistant", tools=tools)

Session and tenant isolation

Bind a session_id (and optionally a scope) when you build the tools. The values are fixed at build time and are not exposed to the model, so an agent cannot read or write outside its slice of memory:

toolset = SpectronToolset(config=config, session_id="user-123")

Two agents built with the same session_id share memory; agents on different session ids stay isolated. See examples/sessionized_memory.py.

Choosing which verbs to expose

By default all of the verbs below are available. Pass include=[...] to expose a subset, for example a collector agent that can only write and a researcher agent that can only read (see examples/multi_agent.py):

collector = SpectronToolset(config=config, include=["remember"])
researcher = SpectronToolset(config=config, include=["recall", "reflect"])

Tools

Every tool returns a JSON-safe dict with a status key that is "success" or "error". A failed Spectron request becomes {"status": "error", "message": ..., "status_code": ..., "trace_id": ...} so the model sees it as data rather than the agent turn failing.

Examples

• examples/quickstart.py - one agent that stores a fact and recalls it.
• examples/sessionized_memory.py - memory isolation across sessions.
• examples/multi_agent.py - a writer agent and a reader agent sharing a context.

Fill in .env from .env.example, then run any example, for example python examples/quickstart.py.

Development

pip install -e ".[dev]"
pytest

The tests use a fake client, so they need no network access or credentials.

Links

• Spectron: https://surrealdb.com/platform/spectron
• Spectron docs: https://surrealdb.com/docs/learn/spectron
• Google ADK: https://github.com/google/adk-python

License

Apache 2.0. See LICENSE.