Memex is the docs and task management system for agents and human teams to ship quality code at AI speed. CLAUDE.md, cursor rules, skills files, ADRs, Notion, Confluence — they were all the right call once. At AI velocity, they go stale faster than anyone can maintain them. With Memex, agents follow the plan, capture every decision and update the plan the moment it’s made, link it to the work it affects — and the agent that picks up the next task with the current context, not yesterday’s document. Quality at AI speed, not vibe coding to a doc nobody’s updated since Q2.
We’ll be in touch when Memex is ready for your team.
We’ll be in touch. Thanks for the interest.
Your CLAUDE.md was right in Q2. Today it steers every agent into yesterday’s architecture. 12 ADRs in the repo while agents made 47 architectural calls last week alone. The wiki with the testing approach was archived three quarters ago — and every agent today still does the opposite.
Every static markdown context layer your team has tried — CLAUDE.md, cursor rules, skills files, ADRs, Notion, Confluence, Google Docs. Agents follow them anyway. Vibe coding to yesterday’s docs.
One strategy doc holds the whole plan — the objective, the architecture, the open decisions and the work items. Every decision gets captured automatically, linked to the work it affects, and injected into the agent’s context the moment it claims that task. The whole team ships against the same plan.
Start in minutes. Works with the tools your team already uses.
One command connects Memex to your tools. Works with Cursor, Replit, Claude Code, ChatGPT and Slack. Point Memex at your repo. The indexer runs in your own infrastructure — self-hosted, fully private.
Build a strategy — the objective, the architectural vision, the open decisions. Decisions are made and captured directly in Memex, your IDE, Claude or GPT. Open decisions notify the right people and are resolved before agents start building. Work items flow from confirmed decisions, each scoped with exactly what an agent needs.
Agents execute one task at a time against the plan, triggered from your IDE. Every decision is recorded. Work is validated against the plan automatically, flagging issues before QA. Conventions and docs update when the code changes. The next agent starts from the last decision made, automatically.
Not a folder. Not a wiki. A complete system.
Memex indexes your repo continuously, so every plan is grounded in the architecture and conventions that actually exist today — not the ones a wiki described last quarter.
Work items aren’t bullet lists. They’re fully scoped briefs, with the relevant decisions, conventions and constraints already attached. The agent picks up the task with the context it needs to ship.
Decisions are captured at the moment they’re made — in planning, in chat, in PR review — and propagate to every task and agent that depends on them. No re-litigation. No drift.
Conventions, ADRs and runbooks update when the code does. Memex keeps the docs and the codebase in lockstep, so what an agent reads tomorrow reflects what shipped today.
Managed by us. Start building in minutes.
Your infrastructure. Your data. Your LLM keys.
Docker + PostgreSQL 15+
Mindset AI is the agentic platform powering in-app AI agents at a growing list of B2B SaaS companies. Every engineer on our team works with coding agents daily. Our product people design in conversation with AI. Our support team uses AI to triage and respond. We aren't just building AI agents — we're running an AI-native company.
When we started, we assumed the hardest part of AI-assisted engineering would be code quality. It wasn't. Modern coding agents produce good code when they know what they're trying to build.
The hard part turned out to be context. Two engineers, each with their own agent, arriving at contradictory decisions. An agent loading a wiki that was last updated two quarters ago. A re-prioritisation that silently invalidates half the decisions that shaped the previous plan.
We run Mindset on a set of mental models that take bottlenecks seriously. One of them is the Theorem of Constraints: any interconnected process operates at the speed of its slowest part. We're allergic to bottlenecks in our own organisation, and we look for them constantly.
When we applied that lens to AI-assisted development, the bottleneck was unmistakable: it wasn't writing code, it was capturing, propagating, and maintaining decisions across humans and agents.
Memex AI is the system we built to remove that bottleneck. It started as internal infrastructure. We use it every day. It's now the decision layer behind everything Mindset builds — and Memex itself runs on the Mindset agentic platform, so the product you're looking at is a working example of what both can do together.
We’ll be in touch. Thanks for the interest.
Free to start. Self-hosted. No credit card required.
CLAUDE.md and cursor rules are static — they reflect what was true when someone last edited them. Memex updates automatically as decisions are made and work progresses, so every agent reads the current state, not a snapshot from three months ago.
No. Memex connects to the tools your team already uses — Slack, your IDE, your repo. Decisions and context get captured automatically from where they happen. You don’t have to write anything new.
The indexer runs in your own infrastructure. Nothing leaves your environment. Memex is self-hosted and fully private.
Yes. Memex connects to Cursor, Replit, Claude Code, ChatGPT and Slack. One command to set up.
Memex checks dependencies across strategies before any agent claims a task — so two agents can’t pick up conflicting work. When a decision changes, Memex automatically marks everything downstream that needs rechecking.
Memex integrates with Linear, Jira, and Notion — it sits underneath them as the decision and context layer they were missing. It doesn’t replace them.