Asking for a local open-source web automation agent: Playwright + local LLM architecture sketch
Weary-Release3436 · reddit · 2026-08-18
The poster wants to automate repetitive browser tasks across arbitrary websites (TMDb metadata, YouTube Studio, bookmark management, cross-site data entry). Tools like ChatGPT Work and Google Antigravity handle this surprisingly well but are cloud services with quotas and limited control.
They sketch a full self-build architecture: Playwright as the browser layer (real Chromium, persistent logins, multiple accounts) with a local small model as the brain deciding only "which tool next, which field matches, did the last step succeed" — suggesting on-device agentic models like Liquid AI's LFM2.5-2.6B may suffice. Planned features include natural-language reusable templates, human handoff for 2FA/verification, a review mode before large changes, and undo/history. The post asks whether such an open-source project already exists and whether the architecture makes sense.
More from coding & agent
- MUZIM workflow: local transcription and exact quote retrieval for bug tracking — CodeByPoonam · 2026-08-19
- Using AI Agents to draft release reports from evidence collections — CodeByPoonam · 2026-08-19
- Netlify integrates OpenRouter to enable model swapping without code changes — thisiskp_ · 2026-08-19
- Dev runs three Codex accounts plus Claude to parallelize coding agents — ChanceKelch · 2026-08-19
- DeepSeek open sources 'deepseek-harness' agent framework with 130k+ stars — alex_verem · 2026-08-19
- Cloud agents save hours vs local setups: a dev's parallel-workflow experience — brandon_galang · 2026-08-19