Diff approval
Before AI changes your files, review the diff and accept or reject each change. Critical assets require approval.
MKPost runs topics, drafts, rewrites, publishing and review locally, and shows a diff before changing your files.
It is not a one-click generator. It is an approvable, reusable desktop workspace for long-term content assets.
Core features
MKPost turns creation into approvable actions: AI moves the work forward, you make the calls.
Before AI changes your files, review the diff and accept or reject each change. Critical assets require approval.
Topics, drafting, polishing, publishing and review each have a dedicated skill instead of one overloaded prompt.
Drafts, materials and sessions stay on your machine. You configure your own model provider, without our proxy.
Move final drafts into published state, create snapshots and ledgers, and use browser automation for platform logins.
Workflow
Run the full content cycle in one line instead of moving work across scattered tools.
Save ideas, links, images and documents into one local workspace first.
Ask the agent to run skills for topics, drafts and rewrites, with confirmation at every step.
Turn the final draft into publish snapshots, channel ledgers and asset records.
Bring results back into the workspace so style, judgments and materials keep compounding.
Pricing plans
For creators who write continuously, need multiple local workspaces and want AI work to become long-term assets.
You pay your own model API costs. We do not proxy model calls.
FAQ
These answers follow the current project docs and pricing strategy. Model usage, cloud drafts and team collaboration are not promised.
No. MKPost uses BYOK. You need to configure your own model provider and API key in the desktop app.
Free allows 1 local workspace. Monthly does not limit local workspace count, so you can manage accounts, columns and projects separately.
The website handles accounts, payment, email verification and desktop auth callback. Drafts, materials, sessions and publish assets stay in local desktop workspaces.
After login, the desktop app opens the auth callback page. The website creates a token and returns to the app through mkpost://auth.
Start a local workspace with your own model key. Let AI move the work forward, while you make the final call.