VK → Telegram + MAX Autoposter
VK posts appear in Telegram and MAX within seconds, with no duplicates
A service that mirrors VK group posts into Telegram and MAX automatically.
Client review
"Thanks to Pavel, the task is done. Always reachable, gave me detailed instructions and a guide, I will come back and I recommend him to everyone."
Before
The owner posted to VK first, then re-uploaded everything to Telegram and MAX by hand. Posts went out late, attachments got lost, and ready-made tools either did not support MAX or could not be trusted with media.
What I built
A service on its own server watches the community wall and republishes each new post to a Telegram channel and a MAX chat right away: text, photos, videos and links. It remembers what it sent and retries only the send that failed.
After
New posts appear in Telegram and MAX within seconds, with no manual work, misses or duplicates. One command brings the service up, and a guide in Russian lets the owner run it without technical skills.
What the client got
-
One post, three platforms
Post on VK, and the Telegram channel and the MAX chat get the same post on their own.
-
Text, photos, videos and links
Albums and videos arrive whole: the media is uploaded separately for each platform.
-
No duplicates, no misses
Every post is logged, and if one platform fails, the retry goes only there.
-
Pinned and old posts
Pinned posts are skipped, and the old feed can be left behind on the first run.
-
Runs unattended
The service lives on its own server, shows that it is running and records what happened if something goes wrong.
-
A guide for the owner
Instructions in Russian: the service can be run without technical skills.
How it looks
How it was built: details and stack
Tech Stack
Overview
VK → Telegram + MAX Autoposter is a small, reliable backend service that watches a VK community wall and republishes each new post to a Telegram channel and a MAX chat. It handles text, photo albums, videos and links, and runs unattended on a VPS via Docker.
Problem
Cross-posting the same content to several social networks is exactly the kind of repetitive work people stop doing well. The client published to VK first, then re-uploaded everything by hand to Telegram and MAX. Posts arrived late, attachments got dropped, and some updates never made it across at all. Off-the-shelf bridges either didn’t support MAX or couldn’t be trusted with media.
Solution
The service subscribes to VK’s Callback API and reacts to the wall_post_new event the moment a post appears - no slow polling. Each post is normalized into text plus a set of attachments, then fanned out to Telegram and MAX through their respective Bot APIs. A SQLite layer records every post so nothing is published twice, and a partial status lets the service retry only the network that failed instead of duplicating the whole post.
Features
- VK Callback API mode (
wall_post_new) with a polling fallback viawall.get - Publishes text, photo albums, videos and links to Telegram and MAX
- SQLite deduplication with
partialstatus and automatic re-delivery on failure - Skips pinned posts and, on first run, optionally skips the existing backlog
- Structured logging (pino) and a
/healthendpoint for monitoring - Docker + docker-compose deployment with GitHub Actions CI/CD to a VPS
Development Process
The trickiest part was media: VK, Telegram and MAX each model attachments differently, and MAX needs files uploaded before a message can reference them. The service downloads media to a working directory, uploads it per platform, and retries transient attachment.not.ready responses. Region also mattered - Telegram’s Bot API isn’t reliably reachable from Russian VPS providers, so the deployment targets a VPS outside RF.
Results
- New VK posts appear in Telegram and MAX within seconds, hands-free
- Deduplication and per-network retries eliminated both misses and double-posts
- Fully containerized: a single
docker compose upbrings the service back on a fresh server - A client-facing Russian guide made the system operable by a non-technical owner
What Was Learned
Reliability in a small integration comes from the boring parts: idempotency, partial-failure handling and honest logging. Modeling each post as “publish to N targets, track each independently” turned a fragile mirror into a service that quietly does its job.
Services used in this project
Related case studies
- Telegram Sales-Funnel Bot for Online CoursesA Telegram bot that runs the whole funnel - lead magnet → guide → intensive → course - with checkout, auto-delivery and an admin panel.
- TikTok Auto-Scheduler for Recipe VideosA tool that matches finished recipe videos to their captions and schedules them to TikTok one a day through the PostMyPost service.