WorkplaneWorkplane
How it worksFAQDocs

The workspace for everything
your AI agents create.

Every document, prototype, and spec stays in one place,
versioned, and ready to share.

Every version savedNothing overwrittenNothing lost

See your agent publish a document in under 15 seconds.

workplane.co/mike/brand-voice
Workplane/

brand-voice

M
By Mike
·Updated 3w ago·

How we write. Short, confident, skeptical of hype. Built for agents to read and humans to skim.

Versions

Current
Files:no changes
Saved versions
Earlier this month
v2Mike · 3w ago
No description
v1Mike · 4w ago
No description
Works with the agents you already use.
Claude CodeCodexCursorClaude DesktopMCP

Using your agent starts with hunting for the right docs

Digging through folders, opening “spec-Final-v2.md”, copying links, @-tagging files — every new chat, you stitch together the context your agent needs.

~/Documents/brand-launch
  • brief.mdMar 14
  • brief-v2.mdMar 19
  • brief-final.mdApr 2
  • spec.mdApr 4
  • spec-v2.mdApr 9
  • spec-final.mdApr 14
  • spec-Final-v2.mdApr 21
  • spec-final-FINAL.mdyesterday
  • brand-voice.mdApr 12
  • brand-voice-v3.mdtoday

“Your agents are only as useful as the context they can reach.”

D
@Deluvio
ghGitHub Issue #39195·Mar 2026
"There is no way to share memory across a subset of projects. A lot of knowledge is relevant to multiple (but not all) projects."
★ openenhancementmemory
KP
Katie Parrott
EEvery.to·Nov 2025
"If I want team members to use a skill, I have to notify the team, send the files, and hope the upload goes smoothly."
31 likes7 comments
N
Neilos
<>DEV Community·Mar 2026
"You're copy-pasting context between sessions, manually tracking which PR depends on which, and babysitting agents that can't see the full picture."
1.1k readHow I Manage 15+ Repos
TS
Titus Soporan
✎The Spine Pattern·Jan 2026
"The 'cold start' problem gets exponentially worse across multiple repositories. Context fragmentation is brutal without some orchestration layer."
long-formmulti-repo
KA
@karpathy
𝕏Apr 2026·1.2M views
"The memory here is a simple collection of files in universal formats… any kind of interface can be used to view them as the output."
8.7k ♥796 ↻
HN
@heynavtoor
𝕏Apr 2026·380k views
"Your explorations compound in the knowledge base. Nothing disappears into chat history."
3k ♥365 ↻
D
@Deluvio
ghGitHub Issue #39195·Mar 2026
"There is no way to share memory across a subset of projects. A lot of knowledge is relevant to multiple (but not all) projects."
★ openenhancementmemory
KP
Katie Parrott
EEvery.to·Nov 2025
"If I want team members to use a skill, I have to notify the team, send the files, and hope the upload goes smoothly."
31 likes7 comments
N
Neilos
<>DEV Community·Mar 2026
"You're copy-pasting context between sessions, manually tracking which PR depends on which, and babysitting agents that can't see the full picture."
1.1k readHow I Manage 15+ Repos
TS
Titus Soporan
✎The Spine Pattern·Jan 2026
"The 'cold start' problem gets exponentially worse across multiple repositories. Context fragmentation is brutal without some orchestration layer."
long-formmulti-repo
KA
@karpathy
𝕏Apr 2026·1.2M views
"The memory here is a simple collection of files in universal formats… any kind of interface can be used to view them as the output."
8.7k ♥796 ↻
HN
@heynavtoor
𝕏Apr 2026·380k views
"Your explorations compound in the knowledge base. Nothing disappears into chat history."
3k ♥365 ↻

Give your agent a place to save its work
so nothing disappears.

Your agent’s research, specs, and prototypes stay organized, versioned, and reachable from any agent or machine.

01

Go back to any version

Each save creates a new version, so you can return to the one that worked whenever you need it.

v4 · live
v3 · last worked
02

Keep related work together

Group your agents’ documents, prototypes, and outputs around the thing you’re building, with a name, description, current version, and history.

brand-launchv3 · 4 files
spec.md
hero.html
logo.png
03

Work on it from any agent

Every grouped work item has a stable link your agents can read from and write back to, across sessions, repos, and machines.

workplane.co/mike/brand-launchv3
Claude Codewrites
Cursorreads
Codexreads & writes
How it works

Turn your agent's work into live, versioned pages.

Your agent creates specs, research, and prototypes scattered across folders and chats. Send them to Workplane. They become live, versioned pages with a URL, reusable in any session.

01 · SAVE

Save what your agent just made

Send files and folders you want to keep to Workplane. You get back one URL with everything inside.

Claude Code
You
save my brand-voice folder to workplane
Claude
Saving 5 files to Workplane...
✓ saved @mike/brand-voice v3
→ workplane.co/mike/brand-voice
02 · KEEP

Send updates, keep every version

Each update creates a new version at the same URL. You can always roll back to an earlier version.

workplane.co/mike/brand-voice
brand-voice v3
5 files · pushed 2m ago
Versions
v3 · draft
v2 · 2d ago
v1 · 5d ago
03 · USE

Use it in any agent

Type the name or paste the URL. Your agent loads the latest version, ready to keep working.

CClaude Desktop
Write a product update using @mike/brand-voice
↳ pulled @mike/brand-voice (v3) · 5 files
Drafting in your voice now. Opening with a short, confident sentence and no hype words...

Share your work. Get feedback.
Keep iterating.

01 · VIEW

Open the URL. Read it like a webpage.

Markdown, HTML, and images all render. Specs, prototypes, and assets become pages — not raw files.

workplane.co/mike/brand-voice
index.mdhero.htmllogo.png
Brand voice v3
A short guide to how we write — for humans and agents.
Principles
  • Short, confident sentences.
  • No hype. No filler words.
  • Show, don't tell.
# avoid
"seamlessly unlock"
logo.png
02 · COMMENT

Comments pinned to specific versions.

Your reviewer leaves a note on v3. The thread stays on v3 even after v4 ships. No lost context.

workplane.co/mike/brand-voiceCopy
S
@sarapinned to v3
"love this — the 'show, don't tell' line is it."
J
@julespinned to v2
"use this framing for the welcome email too"
03 · ITERATE

Iterate without losing what worked.

Push v1 from terminal. Your agent drafts v2. Publish v3. Every version keeps its own URL — roll back any time.

product-spec.mdv1
pushed by cli · Mon
product-spec.mdv2
edited by Claude · Wed
product-spec.mdv3
published · Fri
roll back to v1
04 · CONTINUE

Pick up where you left off — any agent, any session.

Friday's chat closes. Monday morning, the URL loads the latest into your next session, exactly where you stopped.

Claude Code
Mon session
Cursor
Wed session
Codex
Fri session
brand-voicev3
workplane.co/mike/brand-voice
5 files · live

Questions.

Who can see what I push?
Private by default. Only people you share the link with can access it. You control visibility per artifact.
Does this work with Claude Code or Cursor?
Yes. Workplane has agent integrations. Your agent reads and writes directly — no copy-paste.
How is this different from Notion?
Notion isn't built for agents or files. Workplane is — your agent reads and writes directly, everything is versioned, and it renders as a live page.
What is an artifact (and why not just use folders or docs)?
Agents write fast and can get messy, so an artifact lets you and your agent update files safely. An artifact is a versioned folder with files — a spec, a research doc, a playbook — that you view live and roll back if something breaks.
What if my agent makes a mistake?
Ask your agent to save a version anytime. Each version is a permanent snapshot — you can always roll back.
Is it free?
Free during early access. Pricing for teams comes later.

Start with your agents.

Free during early access.

WorkplaneWorkplane
DocsGitHubPricing
© 2026