Lindy
The lowest-friction install for non-technical operators. About 3 minutes to first running Lindy. Plain-English workflow definition; Lindy auto-suggests LUU tools.
What you'll get
Lindy is one of the leading "prompt-to-agent" platforms — typical user is a non-technical operator who describes what they want in plain English and Lindy builds a multi-step workflow with 4–5K integrations. Lindy added native MCP client support in early 2026.
This is the lowest-friction way to put LUU's 27 tools in front of a non-technical end user.
Prerequisites
- Lindy account ($49–$99/mo tier with MCP support)
- An LUU API key — sign up (free tier: 25 calls/day)
Step 1 — add LUU as a custom integration
In Lindy: Settings → Integrations → MCP Servers → Add Custom MCP.
| Field | Value |
|---|---|
| Server name | LUU |
| Description | Strategic-judgment substrate: forecasts, signals, briefs, risk, scenarios, ICD 203 note review, decision-frame |
| URL | https://api.leadershipunderuncertainty.org/api/mcp |
| Auth | Bearer Token |
| Token | your LUU API key |
Save.
Step 2 — create a Lindy that uses LUU
Click New Lindy → describe what you want in plain English. For example:
"Every weekday at 7 AM, fetch today's strategic intelligence brief from LUU, identify any high-priority signals about the EU AI Act or supply chain, and post a summary to #strategy-daily on Slack."
Lindy will:
- Recognize the LUU MCP server is connected
- Suggest
luu_briefandluu_signalsas the right tool calls - Pre-build the schedule trigger and the Slack post step
- Show you the workflow as a visual graph
You confirm or edit. Lindy runs it.
Step 3 — try it on a one-off question
Even without scheduling, you can drop into a Lindy chat and say:
"Use LUU to forecast the probability that TSMC's Arizona fab hits N3 commercial production by end of 2027. Compare scaffolded mode to ensemble mode."
Lindy invokes luu_forecast twice with different mode parameters, then renders both results side-by-side.
Useful Lindy patterns
| Pattern | What you say to Lindy | LUU tools |
|---|---|---|
| Daily brief in Slack | "Every morning post LUU's brief to #intel-daily, highlight act-priority signals" | luu_brief, luu_signals |
| Memo critic on Gmail | "When I forward an email tagged 'review-memo' to Lindy, run LUU's note review and reply with the scorecard" | luu_note_review |
| Source check on inbound RSS | "For every Reuters AI-policy item, run LUU's fact-check and append to my Notion DB" | luu_fact_check |
| Pre-meeting prep | "30 min before any 'Strategy Review' calendar event, generate a perspective pack for the agenda topic" | luu_brief (perspective-pack mode) |
| Risk-spike alert | "When a LUU prediction probability moves more than 10pp, ping me on iMessage" | luu_predictions (poll) or LUU webhook → Lindy |
Why Lindy
The pitch for Lindy specifically: non-technical users build the workflow themselves. You don't need a developer to wire up "monitor this RSS feed → call LUU → write to Notion." A strategic analyst writes one sentence describing what they want, Lindy assembles the steps, and the workflow runs.
This is the lowest-cost-to-customer way to expose LUU as a customizable surface.
Troubleshooting
"Custom MCP" not visible in Settings: Your Lindy plan doesn't include MCP. Upgrade to a tier that does.
Lindy won't auto-suggest LUU tools: Make sure the integration shows ✅ in Settings → Integrations. If yes, be more explicit in your prompt: "use the LUU integration's luu_brief tool" rather than just "fetch the brief."
Tool calls fail intermittently: LUU's MCP rate limits apply per-key. If a Lindy runs many parallel calls, you may hit per-minute limits. Stagger workflow steps in Lindy's editor or request a higher-tier API key.
Next steps
- For more developer control with the same UX, try n8n — also visual but self-hostable
- For programmatic integration in code, see Claude Code reference templates