Meko For Claude

Claude remembers this conversation, but not the decisions from three sessions ago. It can’t tell you what the agent that reviewed your pull request worked out.

Over MCP, Meko provides Claude 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.

Context that lasts over time

The information Claude learned is preserved in your Meko datapack for your next session. It’s accessible to other permitted AI agents and can be shared with your team when required.

Connect as a custom connector for on-demand tool calls, or run the installer in Claude Code for automatic capture with skill and session hooks.

How Meko Benefits Claude

Connect once, then ask in plain language. With a connector, mention Meko in the ask so the agent calls the tools. With the Claude Code installer, the skill automatically calls Meko.

Pick up where you left off

“What does Meko remember about the auth refactor?”

Claude searches your memory before answering, instead of making you re-explain.

Continue another agent’s work

“Ask Meko what my Cursor session figured out about the flaky test.”

Your other agents (Cursor, Codex, a custom one) write to the same datapack, and Claude reads your memories across all of them. Context from a teammate’s agents can be promoted to the shared knowledge base.

Save a decision as it’s made

“Remember in Meko that we’re going with the single-table design, and why.”

Stored with its reasoning, linked to the related context.

Share a conclusion with your team

“Promote that decision to the team project knowledge.”

Memory starts private. Promotion is controlled by datapack owners and maintainers. The shared knowledge base contains only what someone chose to publish.

Store generated artifacts during conversations

“Save that migration script to my Meko datapack.”

The files Claude produces (reports, scripts, diagrams) are stored in your datapack next to the conversation that produced them. They are retrievable in later sessions, by any agent you connect.

Connecting Meko in Claude

Meko is a remote MCP server. Claude connects to it as a connector, with no download, config file, or local install. The connection is made from Anthropic’s cloud, so it works the same in claude.ai, Claude Desktop, Cowork, and mobile.

Adding Meko as a Claude Connector

claude.ai/directory/connectors/meko

Meko is available via OAuth login in the Claude Connector directory.

  • Go to Settings → Connectors.
  • Click Add, then Browse connectors.
  • Type “Meko” then click Add.
  • Click Connect and complete the Meko sign-in. Claude never sees your password, and you can disconnect at any time.

Team and Enterprise Plan Instructions

An Owner adds Meko once under Organization settings → Connectors. Everyone else connects their own Meko account, so each person only reaches the datapacks they already have access to.

Claude Connectors (Meko)

Claude Code

Claude Code has its own installer. It wires session hooks and installs the Meko skill, so your conversations are saved automatically as you work, always on:

npx @yugabytedb/meko-mcp --client claude-code

It talks to the same server at https://mcp.mekodatastg.wpenginepowered.com/mcp, which exposes tools covering memory, conversations, knowledge base, datapacks, artifacts, and observability.

Automatic capture writes to your pinned datapack, or your default datapack if none is pinned. A memory written from Claude Code is immediately readable in Claude Desktop.