Project memory for
humans and AI agents
Logiqo is not just a code map. It is the living record around your project: plans, deadlines, research, documentation, competitor notes, measurements — all connected, so a year later you find what you chose and why, instead of guessing again.
Sound familiar?
Claude and Cursor don't know your project — what's already built, why it was done this way, where you're heading.
Your docs either don't exist or lie a month after someone wrote them.
And next week you're onboarding someone new — explaining it all again, by hand.
Who it's for
One shared memory instead of tribal knowledge — for every teammate and every agent they bring.
No more permanent first days: Claude and Cursor know the project's why, not just its files. Start free — your graph is ready for the team when they are.
Your agents read and write it — over MCP
Claude Code working with a Logiqo project over MCP — reading what's there, recording what changed.
The timeline
The timeline is where a person works: what is due, how much of it, by when, and who is on it.
And on the same axis, behind today, sits everything that already happened — when a decision landed, when a piece of code was mapped, when a note changed. Planning forward and history backward on one line, not two tools.
Architect
Agents that work on your code, against the memory you already have.
You describe the goal in your own words. Agents do the work, review each other, and stop at an open pull request. You look at what changed, and you press merge. Nothing lands without you.
It runs on your machine, on your own subscription. We never see your code, your data or your keys.
Vector + full-text search finds the node, then walks its connections, so you get the neighbours of the answer, not just the answer.
Nodes point at exact files and lines, and the graph shows when the code moved on and a note went stale.
Every agent write queues as a draft with a diff. Approve, reject, or fix it first.
One URL. No local process, no config file. Sign in in the browser and pick which project Claude or Cursor sees.
Pricing
Start free. Pay for what you actually use.
- First document import free — up to 160K characters
- 30 IQ welcome credit to explore chat and the graph
- 1 project, up to 300 nodes
- MCP agent access — 30 calls/month
- 400 IQ monthly, accumulates up to 800
- Up to 5 projects, 10,000 nodes
- MCP agent access — read + write queue
- 1500 IQ monthly, up to 3000
- Unlimited projects, up to 50,000 nodes
- MCP agent access — higher limits, multi-project
- Everything in Pro, for you
- 3 seats for invited members (you are not one of them)
- Per-bundle access control for every member
- Additional seats $40/mo each
- Bought per project
- Larger teams and custom volume
- Security review and custom terms
- Onboarding and direct support
- Self-hosting and data residency — let's discuss
Fair usage-based credits: every action costs IQ proportional to its real compute cost — light questions cost less, heavy imports cost more. No surprise bills, no upgrade walls.
Founding pricing — your rate is locked in for the life of your subscription.
FAQ
Read your graph — search it, expand nodes, follow anchors down to the exact lines in your repo. And propose changes: new nodes, edits, links. Everything it writes queues as a draft for you to approve or reject.
Those store what the agent said. Logiqo stores what your project is — a curated map you own, that outlives the model, the client and the chat. Claude, Cursor, Codex and any new developer's agent all read the same map. Nothing enters it without your approval.
Sign-in is OAuth 2.1 — you approve the connection in your browser and choose exactly which project the agent can reach. The token is bound to that project and nothing else. API keys remain available for clients that need a config file. And agents never overwrite approved knowledge: everything they write is a draft.
There's a chat inside the app. It's a convenience, not the point — the graph is built to be read by Claude and Cursor.
Logiqo's usage credit. Each AI action costs IQ proportional to its real compute cost. Browsing your graph is free.
Documents and text — Logiqo splits them into linked nodes automatically.
Security
- Your data is hosted on Supabase, encrypted in transit (HTTPS) and at rest.
- Row-Level Security isolates every project's data from every other.
- Agents sign in with OAuth 2.1 — the token is bound to one project and nothing else. API keys remain available for clients that need a config file; they are stored hashed, never in plaintext, and scoped to a single project.
- We never train any AI model on your data, and neither does our AI provider, Anthropic.
- Security contact: security@logiqo.org.