Dots dashboard experiment: hand-rolled scrapers, no Firecrawl, flaky schedules
edwin · x · 2026-10-03
Part 6/7 of edwin's retrospective on having OpenAI's Dots build an AI status dashboard 8 times:
- Scraping: all hand-rolled HTTP with custom JSON, RSS, and HTML parsers — no Firecrawl, Browserbase, or Tavily
- Scheduling: some builds shipped with no active schedule at all; where one existed it used a ChatGPT task rather than a Cloudflare cron
Takeaway: agents still reinvent wheels and remain unreliable at production-grade scheduling and parsing — useful signal for agent engineering.
More from coding & agent
- Turso joins Supabase to build the database platform for the agentic era — edwin · 2026-10-03
- Service size limits shift from two-pizza teams to one architect's head — corbtt · 2026-10-03
- 5 decision models tested on 1,000 real transactions — forcing yes/no beats abstention — purealgo · 2026-10-03
- AI coding agents' most useful trait: they can hold still without complaining — tdhopper · 2026-10-03
- Prediction: Claude/Codex Will Eventually Morph Into Slack — andreisavu · 2026-10-03
- One-Shotting Five Native macOS Apps in a Week With Opus 5.5 — mneary0 · 2026-10-03