Sarmad Sandeelo — Troubleshooting, Client Docs & Programmatic Clay Tables
Using Claude Code to troubleshoot and generate client documentation — plus a bonus: building Clay tables programmatically. (Extension attached in the original post.)
- Watch the loom below (2x speed is fine).
- Comment at the bottom of this page — what landed, what you'd steal, what you'd do differently.
- If this is the use case you're trying this week, record your own loom of it and drop it in the channel.
The deepest build of the week, and a continuation of Sarmad's "repeatable systems" thread. It runs on a Clay Chrome extension he built — LSClayStack — that exports any Clay table's JSON so Claude can work with it. Three use cases worth calling out, all on the CallRail PE project: (1) Troubleshooting — point Claude Code at the exported table JSON and it maps the whole workflow (which tables feed which) and audits the logic; in the demo it caught an AI enrichment column with no conditional run firing on every contact and burning credits, before it ran on the full table. (2) Client docs — feed the JSON + project context and it generates the documentation, then converts MD → a clean PDF via the LeanScale Report Generator skill. (3) Building new tables programmatically — generate the table JSON and import it via the extension. The key gotcha to highlight: Clay enrichment "action package IDs" are unique per workspace, so you must feed the client's own table JSON first — you can't reuse one JSON across clients.
Discussion
Sign in to comment on this lesson.