Most OpenClaw setups are context-blind. The agent completes the task meaning it runs the cron, scrapes the page, posts the summary with no idea whether you slept four hours or eight, whether you're in back-to-back meetings or have a rare clear afternoon, whether your body is signaling recovery or overreach. The automation works but it’s not a clear view of reality.
These six recipes fix that. Each one connects OpenClaw to Fulcra, which provides your agent with real world knowledge about your calendar, sleep, recovery, HRV, wearable data and then passes it to the agent as context before anything runs. The result is that you get the same automations you'd build anyway, except they adapt to you.
These are ideas for skills. You'll get to build the actual skill.
Ten minutes before any calendar event, the agent fires.
It pulls attendees, your last interaction with each of them (email thread, Slack message, prior meeting notes), current location vs. meeting location, and a one-liner from Fulcra:
Posts to Slack, WhatsApp, or any channel you configure. Arrives when you need it, not when you think to check.
Most workout-planning automations are single-wearable. The real problem with stitching them together isn't finding an MCP server for each device, those exist. It's that each one reports data differently. Your Oura ring's recovery score is not the same scale as your Whoop's. Your Garmin's training load calculation doesn't map cleanly to either. The agent ends up writing normalization code or running math tools to make sense of it all, which costs tokens and time on every run.
Fulcra pre-normalizes recovery data across Whoop, Oura, Garmin, and others into a consistent format. The agent reads a your 7-day training load and tomorrow's calendar density, then proposes a workout, or a rest day, that fits both your body and your schedule:
"Recovery is 72/100, last hard session was 3 days ago, you have a 45-minute gap at 6pm. Recommendation: moderate run, 30–35 minutes, zone 2."
"Recovery is 41/100, you have a high-stakes presentation tomorrow morning. Rest or a short walk only."
Fulcra detects a flight in tomorrow's calendar. The night before, the agent assembles:
Posts to WhatsApp the night before in one message. This is everything you'd otherwise reconstruct from four apps at 5am.
A Sunday-evening cron fires. The agent asks Fulcra's tools for the week's data including sleep trends, calendar density, exercise consistency and assembles a review:
This review synthesizes it into a concrete recommendation because the agent has both the state data (from Fulcra) and the schedule context (from your calendar) making is a thinking partner for your schedule, not just a task runner.
Fulcra detects a healthcare appointment in your calendar. The agent assembles a printable summary:
You walk in with actual data instead of "I've been sleeping okay, I think."
All recipes require:
Fulcra was designed by people who get privacy and know the importance of an infrastructure solution that can be the secure private datastore for the rest of your life. Here data is yours, under your control, and only shared with the people and tools you choose to share it with.
