Meko for Strands Agents
Meko gives your Strands agents memory. Your code decides what to remember, when to recall, and what to record. A fact saved today is available tomorrow, for this agent and for every other one that uses your datapack.
Over MCP, Meko provides Strands Agents with:
Agent-scoped memory
What your agent learns persists beyond the session and is readable via every AI tool you connect to.
A shared knowledge base
Documents and promoted conclusions are searchable by all agents and teammates with access to the datapack.
A trace of every read and write
Each conversation is stored with its reasoning, so you can see what an agent did and why.
Your Strands MCP client is the integration point
Strands connects to MCP servers via a built-in client, so a few lines give your agent access to all of Meko’s tools.
Operations covering memory, conversations, knowledge base, datapacks, artifacts, and observability are available for you. Your agent calls them when your code or your prompt says to, and your code can call the same endpoint directly when an operation must happen every time.
How Meko Benefits Strands Agents
Remember in one ask, recall in the next
Tell the agent to remember a decision, and ask for it later in plain language. The answer comes from the datapack. Ask in the same run, in a new run, or from a different agent entirely. Using memory tools, you can read any agent’s memories on the datapack by agent ID.
Search before reasoning
Call the memory and knowledge base search tools at the start of a run, and the agent begins with prior conclusions, uploaded documents, and team knowledge to enrich their context.
Record a run and its files, read them back later
Conversation tools log the steps you choose to record: what the agent was asked, what it did, and why. Artifact tools store the files the run produced next to it. Open both later from any tool on the datapack.
Promote conclusions to team knowledge
Memory starts private to the datapack. Memory tools can promote a conclusion to the shared knowledge base, where every agent and teammate with access can search it.
Give a whole multi-agent system one memory
Strands runs agents in graphs and swarms. Each one writes to the same datapack, tagged by agent. Tool filters control which groups each agent gets: give the researcher the search tools, give the writer the memory tools.
Connecting Meko in Strands Agents
- Install the SDK: pip install strands-agents
- Connect Meko with the built-in MCP client:
from mcp.client.streamable_http import streamablehttp_client from strands import Agent from strands.tools.mcp import MCPClient meko = MCPClient(lambda: streamablehttp_client( "https://mcp.mekodata.ai/mcp", headers={"API_KEY": "<your-api-key>"}, )) with meko: agent = Agent(tools=meko.list_tools_sync()) - Generate the API key in the Meko portal under Settings, then API Keys.
- Run the agent. All of Meko’s tools are available to it, and your code can call the same endpoint directly whenever you want recall or logging guaranteed.
- To see a full run recorded and read back, clone the Strands Agents + Meko sample.