MemClaw
MemClaw is a governed shared memory system for AI agent fleets. It is described as a shared cognition layer for enterprise AI agents, with a focus on letting one agent’s discoveries be recalled by the right teams instead of remaining isolated in separate silos.
The product’s core functions center on multi-agent and multi-fleet recall, governed access, and memory that changes over time. Its features include permissions, audit trails, tenant isolation, visibility scopes, agent trust tiers, row-level isolation, and a full audit log on every operation. It also includes per-agent retrieval tuning, an LLM crystallizer, and lifecycle automation, with the stated aim that memory improves itself through use.
MemClaw is presented for enterprise AI agent deployments and fleet use cases. It is delivered as an online database and offers two integration paths: MCP for any AI client and an OpenClaw plugin for fleet deployments. The page also shows an MCP connection example for Claude Desktop, Claude Code, Cursor, and Windsurf. The service says it is connected, and its status area lists connected tenants, agents, and memories.
Pricing, documentation, and use-case navigation are available on the site. MemClaw is open source under the Apache 2.0 license and references SOC 2. The page also states that it is in production at eToro and notes 31,000 downloads in its first month.
Overview
8 featuresIn the Databases (SQL, NoSQL, vector, graph) space, MemClaw takes a focused approach. It focuses on allowing AI agent fleets to securely share, persist, and govern memory so agents can learn from each other without data leakage. It is built as an open-source project for AI infrastructure engineers and agent developers. The project is open source (Apache-2.0). It ships for API, and it can be self-hosted.
Caura AI builds and maintains MemClaw, and it first shipped in 2026. The project is developed in the open on GitHub with 306 stars and 435 commits in the last 90 days. Key capabilities include shared memory, governed access, and audit trails. It exposes integrations via a public API and an MCP server.
Summary written by a language model from the project’s public pages.
- Shared memory
- Governed access
- Audit trails
- Tenant isolation
- Agent learning
- API integration
- Open source
- Multi-agent support
Built with & integrations
- Google Cloud
- x-cloud-trace-context header
- Next.js
- /_next/static/ in the HTML · __next_f in the HTML
Trust & compliance
Indexing history
1What PulseGate has recorded for this listing
- Indexed11 Jul · 18:20 UTCcaura-ai/caura-memclaw verified against its public sourceSource: PulseGate · Open
Frequently asked questions about MemClaw
- What is MemClaw?
- MemClaw focuses on allowing AI agent fleets to securely share, persist, and govern memory so agents can learn from each other without data leakage. It is catalogued under Databases (SQL, NoSQL, vector, graph) on PulseGate.
- Who should use MemClaw?
- MemClaw is an open-source project built for AI infrastructure engineers and agent developers.
- Is MemClaw free?
- Yes — MemClaw is open source under the Apache-2.0 license and free to use.
- What platforms does MemClaw run on?
- MemClaw runs on API. It can also be self-hosted.
- Is MemClaw still maintained?
- The GitHub repository shows 435 commits in the last 90 days.
- What projects are similar to MemClaw?
- Similar projects tracked by PulseGate include ClawMem, MemClaw, and memclaw.ClawMemMemClawmemclawSee all 15 alternatives
- Who makes MemClaw?
- MemClaw is developed by Caura AI.
- How long has MemClaw been around?
- MemClaw first shipped in 2026.
Similar projects
Closest matches by what these projects do