CrewAI

Expose GoodMem's RAG REST surface as CrewAI tools so any agent can store documents, run semantic search, and manage spaces.

Stable· Python (3.11–3.13)

crewai-goodmem

Overview

GoodMem is a self-hostable RAG service that handles embedding, chunking, storage, and retrieval server-side. The CrewAI integration exposes GoodMem's v1 REST surface as eleven crewai.tools.BaseTool classes, so any CrewAI agent can store documents, run semantic search, and manage memory spaces directly as part of a crew.

Installation

pip install crewai-goodmem
# or
uv add crewai-goodmem

The pip package is crewai-goodmem; the import name is crewai_goodmem. Python 3.11–3.13 is supported, and the only runtime dependencies are crewai and requests.

Configuration

Variable	Description
`GOODMEM_BASE_URL`	Base URL of your GoodMem server (no `/v1` suffix), e.g. `https://your-goodmem-server.example.com`.
`GOODMEM_API_KEY`	API key issued by your GoodMem instance (starts with `gm_`).
`GOODMEM_VERIFY_SSL`	Set to `false` to skip TLS verification for self-signed dev certs (default: `true`).

Each tool also accepts base_url, api_key, and verify_ssl constructor arguments, which take precedence over the environment variables if you prefer to configure them in code.

Quick start

from crewai import Agent, Crew, Task
from crewai_goodmem import GoodMemRetrieveMemoriesTool

retrieve = GoodMemRetrieveMemoriesTool()  # reads GOODMEM_BASE_URL / GOODMEM_API_KEY

researcher = Agent(
    role="Knowledge Researcher",
    goal="Answer questions from the team knowledge base.",
    backstory="You always cite the retrieved memory chunks.",
    tools=[retrieve],
)

task = Task(
    description="Search space '<your-space-id>' for 'agent skills' and answer.",
    expected_output="A grounded answer.",
    agent=researcher,
)

print(Crew(agents=[researcher], tasks=[task]).kickoff())

Local development. The example above uses a TLS-verified production URL. If you are running GoodMem locally with a self-signed certificate, use https://localhost:8080 and disable TLS verification with the client's verify-SSL option (see the Configuration table above). Keep TLS verification enabled for any deployed server.

Retrieval post-processing

GoodMem's retrieval call accepts these optional post-processing parameters (each framework exposes them under its own naming convention):

Parameter	Range	Description
`reranker_id`	UUID	Reranker model that reorders matched chunks by relevance.
`llm_id`	UUID	LLM that generates a contextual answer (`abstractReply`) alongside the chunks.
`relevance_threshold`	0–1	Minimum relevance score for a result to be included.
`llm_temperature`	0–2	Sampling temperature for the LLM post-processor.
`max_results`	integer	Maximum number of results to return.
`chronological_resort`	boolean	Re-sort the final results by creation time instead of relevance.

Learn more

The full README, examples, and API reference are available on the PyPI package page.