CommentForge
Surfaces good tech posts and drafts in-voice LinkedIn comments, sending link plus comment to WhatsApp for manual approval.
- Started
- 2026
- Stack
- Python, Docker, linkedin-api, WhatsApp
- Kind
- Experiment
- Source
- Private
Why I built it
Engaging well on LinkedIn is worth doing by hand, but discovering posts and drafting replies is tedious, and no automation should touch the real account.
What it does
- Surfaces tech posts worth commenting on
- Drafts a human, in-voice comment for each
- Sends link plus comment to a WhatsApp group for approval, with manual paste only
- Runs isolated with its own database, non-root, and no route to other services
How it is built
An isolated Docker service that uses a burner account for read-only discovery via linkedin-api authenticated with a browser cookie, keeping the real account manual-paste only.
How it evolved
- 2026Scoped
Engaging well on LinkedIn is worth doing by hand, but discovering posts and drafting replies is tedious, and no automation should touch the real account.
- 2026Built
An isolated Docker service that uses a burner account for read-only discovery via linkedin-api authenticated with a browser cookie, keeping the real account manual-paste only.
- 2026Shipped
Built as a working prototype.
Built by Syed Tasavour. More at syedtasavour.com.