Dev.to
7/17/2026

The original title is "API-first or browser automation? Lessons from shipping content autoposting"
Original: API-first or browser automation? Lessons from shipping content autoposting
Short summary
A practical engineering guide on building a multi-platform content publishing pipeline, arguing for API-first adapters over browser automation due to fragility and platform bans. The author details per-platform auth and content support (Telegram, Bluesky, Mastodon, DEV.to, Ghost, LinkedIn), shares hard-won lessons on auth state management, native dialog handling, and API version headers, and recommends a one-adapter-per-platform contract pattern.
- •API-first approach: use browser automation only when no API exists, as platforms like X ban scripted site access
- •Per-platform adapter contract: validate, prepare, publish, verify, collectResult — adding a platform means one new file
- •Practical gotchas: build re-auth as automated step, put API version dates in config not constants, assert on DOM absence not presence
Generated with AI, which can make mistakes.
Is this a good recommendation for you?



