For agencies · startups · remote-first teams

Every person on your team
starts every AI session
knowing nothing
about your company.

And no AI platform will ever fix that. Their memory is always individual, always platform-locked — by design. CacheTank builds the shared layer they won't.

Free to start · $8/week for the whole team · Works in any AI tool

Works in Claude, ChatGPT, Grok, Perplexity
Personal context per member
Shared company knowledge URL
The Platform Trap

Why Claude and ChatGPT
will never fix this.

It's not a technology problem. It's an incentive problem. The same platform that holds your AI memory would have to make it work inside their competitor's product. That's never happening.

Every session starts cold

Brand voice, product decisions, client briefs, technical constraints — every team member re-explains it every time they open a new AI chat.

Memory is always platform-locked

Claude's memory stays in Claude. ChatGPT's stays in ChatGPT. The moment someone switches tools — or you change your company's approved platform — it's gone.

No platform will build what you need

A shared team context layer that works in Claude AND ChatGPT AND Perplexity? That would require each platform to support their competitor's product. Their incentive structure prevents it.

The structural truth: Claude's memory and ChatGPT's custom instructions exist for one reason — to keep your context locked inside their platform. Your context is why you stay. CacheTank is the exit.
"Not even Claude or ChatGPT will make that — it would require them to make your knowledge work inside their competitor's product."
The CacheTank Solution

The shared layer no platform will build.

CacheTank gives every team member personal context — who they are, how they work, their expertise — plus a shared company context URL that loads in every AI tool, for every person on the team.

It works in Claude. It works in ChatGPT. It works in whatever tool your team switches to next year. It's not locked anywhere.

Start your team free →
Personal context — per member

Every person's AI already knows them

Each team member gets their own /ctx URL. Role, expertise, working style, preferences. Every AI they open already knows who they are before the first message.

Shared company context — one URL

Every AI already knows your company

Brand voice, product decisions, client briefs, technical constraints — one shared /ctx URL that every person on the team loads in every tool they open.

Who it's for

Built for teams who live and die by their knowledge.

🎯
The Agency

Client context is your inventory

5–20 people. Every client has a brief, a brand voice, a set of constraints. Every AI session starts from scratch.

  • Load client context in two words
  • Every person on the account starts briefed
  • Client voice travels across Claude, ChatGPT, whatever
  • New team member? They're up to speed in minutes
🚀
The Startup Team

The whole company in every AI session

2–10 people building together. Product decisions get made in Slack. Half the team never sees them.

  • Product decisions, brand voice, tech constraints — always loaded
  • Every engineer knows the product vision before they open Cursor
  • No more "can someone send me the context again?"
  • New hire onboards in their first AI session
🌍
The Remote-First Team

Async knowledge that works everywhere

Distributed across timezones. Heavy on Slack and Notion. No shared AI knowledge layer. Until now.

  • Company knowledge in every AI tool, every timezone
  • Async-first teams stop re-explaining in real time
  • Knowledge compounds across the whole team, not just one person
  • Works in whatever tools each person prefers
Not locked anywhere

Your team uses
every tool.
Context travels with them.

Different people use different tools. CacheTank works in all of them — no IT approval, no integrations, no lock-in.

🤖
Claude
Chrome extension + native MCP — loads automatically
💬
ChatGPT
Chrome extension — click to load in any session
🔍
Perplexity
Paste /ctx URL — instantly context-aware
Grok
Chrome extension works in any browser-based AI
💻
Cursor / Windsurf
Native MCP — context in every dev session
🏠
Local Models
Ollama, LM Studio — /ctx URL works anywhere
Whatever AI tool your team uses tomorrow — CacheTank works there too. One URL. No re-setup. No waiting for an integration.
How it works

Set up in 10 minutes.
Works forever.

One admin creates the shared company context. Everyone installs the Chrome extension. Done.

1

Admin saves company context

Brand voice, product decisions, client briefs — say Cache it or click the extension. Takes two seconds per save.

2

Share the company URL

One shared /ctx URL for the whole team. Every member loads it in the AI tools they use. Takes 30 seconds per person.

3

Every AI session starts briefed

From the first message, every AI on the team knows your company — brand, decisions, constraints, current projects.

Personal (per member)
cachetank.com/alex/ctx
Who Alex is, how Alex works
Shared (whole team)
cachetank.com/acme/ctx
What Acme is, what's been decided
The compounding advantage

Company knowledge that grows automatically.

Every decision your team saves goes through a lifecycle — from raw note to guiding principle. The longer you use CacheTank, the smarter your company context gets. No one has to manage it.

Year 5
Institutional knowledge your competitors don't have. Even if your team has completely changed.
0
Level 0 — Raw
Saved
A decision gets made. Someone says "Cache it." Timestamped, filed, available to every AI tool immediately.
1
Level 1 — Confirmed
Acted on
The team referenced it. It proved useful. It earns more weight in the company context bundle.
2
Level 2 — Pattern
A recurring theme
Shows up across multiple saves. The system recognises it as a pattern. It rises in priority.
3
Level 3 — Principle
How you work
Automatically elevated. Every AI on the team knows this as a core rule of how your company operates.
Pricing

The whole team.
$8 a week.

Up to 18 seats. Personal context per member plus shared company context. Weekly billing — take a break anytime.

Team Plan
$8/week
Up to 18 team members · Cancel or pause anytime
  • Personal /ctx URL for every team member
  • Shared company context URL — works in every AI tool
  • Unlimited saves per member
  • Chrome extension + MCP server for every seat
  • Works in Claude, ChatGPT, Grok, Perplexity, local models, anything
  • Wisdom lifecycle — company knowledge compounds automatically
  • 19+ seats? Contact us for enterprise pricing
Start your team free →
Free
$0
30 saves · Personal /ctx URL · No expiry. Try it before you commit the team.
Pro
$2/wk
Unlimited saves · For individual founders and power users.
Start today

Stop briefing
the company
from scratch.

Set up in 10 minutes. Every person on your team starts every AI session already knowing your company.

Free to start · $8/week for the whole team · Works in every AI tool