Maple: the resident automation desk
Maple is my dedicated OpenClaw agent. It keeps the Good to know ecosystem moving by writing scripts, wiring cron jobs, posting WhatsApp updates, and shipping new sections the moment I ask.
What Maple does
- Editorial automations: refreshes the hourly Bitcoin debate, generates Polymarket trading sims every 15 minutes, and spins up info drops like “Where to buy Lunar New Year dishes.”
- Ops glue: creates Node scripts, tags HTML with update markers, schedules cron entries, and tails logs so nothing silently fails.
- Chat updates: sends me WhatsApp summaries whenever a block gets regenerated or a script throws an error.
- Safety rails: refuses medical/legal diagnoses (see the MRI page) and keeps a log of every automated change for quick audits.
Why Maple matters
Good to know behaves like a newsroom + quant desk, but I don’t keep a big backend team. Maple gives me the same responsiveness from a single OpenClaw workspace:
- Speed: ideas go from WhatsApp message → live block in minutes.
- Consistency: every tile uses the same typography tokens and layout since Maple edits the HTML/CSS directly.
- Observability: cron logs live in
/var/log, scripts sit in /home/ubuntu/.openclaw/workspace/scripts/, and Maple’s chat notes link to each run.
- Human in loop: I still approve copy/design; Maple handles the mechanical work, then reports back.
Recent Maple deliveries
- 15-minute Polymarket agent dry-run block (auto-updated via cron).
- t-test vs z-test cheat sheet.
- “AI tools” retrospective written in first-person voice.
- MRI safety disclaimer instead of risky diagnoses.
Ask Maple for…
- Data pulls or dashboards powered by the existing Caddy proxy.
- Static micro-sites (restaurant lists, research notes, project intros).
- Recurring automations: scripts + cron + logs.
- Chat alerts when something changes or breaks.
Just drop instructions via WhatsApp or the OpenClaw console—Maple lives for small, shippable tasks.