A complete LinkedIn content pipeline: discovery, drafting in your voice, validation, scheduling, and publishing. Here's what's live today, and what's on the roadmap.
Finds timely articles and developments related to the subjects that define your practice.
Searches Hacker News, Reddit, GitHub, YouTube, and the web
Scored for relevance and predicted engagement before you ever see them
Grounded in the source
Live
Reads the article itself, so the draft starts from context rather than a headline.
Fetches and reads the full source material
The draft is grounded in what the article actually says
Skipped when you supply your own article instead
Sounds like you, every time
Live
Every post is written in your voice — choose a preset or build a reusable profile from your own writing samples.
Professional, Conversational, or Bold presets
Custom voice profiles trained on your writing samples
Post length adjustable from ~75 to ~375 words
Your point leads the draft
Live
Add your position on a source and the draft is built around it instead of summarising the article.
Takes you reuse are saved for the next post
Set a default take in Voice & Brand
Your take leads the interpretation, not a summary
Six ways to frame the argument
Live
Choose the structure that fits the point instead of accepting one house style.
Source to Insight, Adversarial, Research Lens, Product Promotion, Pattern Recognition, Changed Mind
Selected in the Shape the draft step
The structure shapes the draft; you still edit and approve it
Start from your own expertise
Live
Describe what you already know and PostForge finds the evidence that serves it.
Five kinds of insight: pattern, decision rule, myth vs reality, client question, framework
Each brief names who it affects, where it stops applying, and what to do next
The source supports your point; it does not replace it
Bring your own article
Live
Paste any article URL and PostForge drafts a post from it.
Your blog posts, announcements, or research
Skips discovery and goes straight to drafting
Same quality checks and publishing pipeline
You stay the editor
Live
You see the source, the draft, and the final decision before anything goes live.
Review is the default — approve or edit each draft
Every draft shows its source, checks, and scorecard first
Auto-publish is opt-in and reversible at any time
Catches what reads like AI
Live
Every post passes a repeatable set of checks before it can publish.
Word count and formatting checks
Banned AI-ism filter
Failing drafts are held for review, never auto-published
Check claims against the source
Live
Compare the draft with the article it came from before you approve it.
Every claim is marked supported, contradicted, not established, or insufficient
Quotes the deciding passage when there is one
Included on paid plans — your judgement still makes the call
A scorecard with every draft
Live
See how the draft scores on the qualities that decide whether a post is worth publishing.
Eleven dimensions, from insight potential and evidence quality to promotional risk
Shown before you approve — never part of the published post
Use it to decide whether to edit, redraft, or reject
No repeated sources
Live
Never post about the same source twice.
Article URLs are checked against your entire post history
Matches are filtered out before drafting begins
Applies to discovered articles and your own URLs
Keeps your publishing consistent
Live
Publish daily, weekly, or on weekdays inside the window you choose.
Daily, weekly, or weekday-only schedules
Choose the days and the time window each post goes out in
Timezone auto-detected; pause or change any schedule from Automation
Goes straight to LinkedIn
Live
Posts go live on LinkedIn through the official API — on your schedule.
Direct LinkedIn API publishing (OAuth 2.0)
Draft mode: review and edit before publishing
Auto-publish mode: hands-free after quality checks
Works with your AI assistant
Live
Connect Claude Code or another MCP client and run the same workflow from the tools you already use.
Scoped tokens: full access, draft only, or read-only
Same topics, voice, review, and publishing limits as the app
Create and revoke tokens in Settings → MCP access
Optional AI-generated or uploaded images
Live
An optional enhancement — generate a visual matched to your brand, or upload your own.
Six visual themes (Corporate, Editorial, Bold, and more)
Composition style controls and custom instructions
Upload your own image in Create Post — available on every plan, including trial
Multi-platform publishing
On the roadmap
LinkedIn today. More text-first platforms planned.
LinkedIn: live now
X, Threads, and Bluesky: on the roadmap
Per-platform character limits and norms
Performance analytics
On the roadmap
Impressions and engagement analytics for your posts.
Post performance metrics: on the roadmap
Publishing history available today in the Posts page
What PostForge intentionally doesn't do
PostForge is built for one professional LinkedIn profile, not agencies or teams. It publishes posts; it doesn't write first comments or manage multiple accounts.
Can we use analytics? We use PostHog (our analytics provider in the US) to see howPostForge is used, which sets cookies and sends usage data. Nothing is collected until you accept, and you can change your choice any time in Settings.