Build AI workflows that remember, stay governed, and hand off cleanly.
Start with MemQ, the flagship memory layer for everyday AI users, developer teams, and enterprises. Add Shield for AI security, Legion for approval-ready agent work, and MARS when developers need the open-source runtime foundation underneath the system.
These guides are organized around the work you want to ship: remember context, protect sensitive usage, coordinate agents, prove performance, and expand only when the workflow is ready.
Start with the outcome you need
Make AI remember
Connect MemQ to your AI tools so useful decisions, repo facts, customer context, and team handoffs survive the next session.
Protect sensitive work
Use Shield to reduce risk around prompts, files, browser sessions, and tool actions before AI adoption spreads.
Coordinate agents
Use Legion to keep agents, teammates, marketplace modules, and OpenClaw workflows aligned with shared memory and approvals.
Build the runtime
Use MARS and the MARS SDK to build agents around identity, journal, memory, and handoff primitives in a public framework.
Choose the rollout path
Review setup checks, product pages, and evaluation signals that help teams decide what to deploy first and what to add later.
Product signals buyers and builders can inspect
Use these signals as starting evidence, not marketing filler. Each product page links to the setup notes, operating assumptions, and next steps behind the result.
Broad
Policy coverage
Shield · Coverage spans sensitive data, unsafe prompt patterns, compliance workflows, and team policy review.
35
Memory tools
MemQ · Your AI auto-discovers 35 hosted MCP tools for storing, searching, and recalling memory — no setup required.
4 stages
Runtime loop
MARS · MARS models agent work as perceive, plan, execute, and reflect so behavior stays inspectable.
3
SDK calls
MARS SDK · The initial SDK exposes clearance verification, journal logging, and memory recall.
MemQ
Workflow memory
Legion · Run notes, research, watchlists, and module outputs can carry forward between teammates and agent runs.
Recommended first moves
Pick the path that matches your current bottleneck. The stack works best when memory proves value first, then governance and agents are added to a workflow people already trust.
Set up memory with MemQ
Connect an MCP-capable client, save one useful decision, and confirm the next AI session can recall it without a fresh explanation.
Start with MemQ →Protect AI usage with Shield
Add prompt, file, browser, and tool-action controls before sensitive context becomes part of everyday AI work.
Review Shield →Coordinate agents with Legion
Give agents shared context, approval checkpoints, and a cleaner path for team handoff once memory and governance are working.
Explore Legion →Build on MARS
Clone the public runtime, run the SDK tests, and build agents around clearance, journal, memory, and distributed handoff.
Open MARS docs
