survey

Trigger

Use this skill when the user wants to start building a Gmail classification pipeline, says "set up my inbox", "help me organize my email", "start the Gmail project", or any variant of beginning a new email classification project from scratch.

Purpose

Runs the onboarding interview, catalogs existing Gmail labels (with message counts), samples a small cross-section of emails within those labels, and creates three shared project artifacts in a Google Drive folder: Tracker (XLSX or Google Sheet), Bible.md, and _status.md.

May invoke the handoff skill mid-run if label volume or inbox sampling needs exceed what Claude should handle directly. This is by design — the survey builds the draft-one blueprint only. Task orders sent to subcontractor agents do the heavy lifting.

North Star Constraint

Never walk the full inbox. If the user requests a comprehensive scan of all email, redirect gently: "My job right now is to build the draft-one blueprint. The task orders we send to other tools will dig into the details — that's their job at this stage, not mine."

Skill Stub — Implementation Pending

Full SKILL.md to be written in build session. Load CW-020_Concept.md for scope and design intent.

Key Interview Questions (to be formalized in build)

Where should I create your project folder? (Google Drive — ask for a folder name or parent)
Do all the tools you plan to use have direct access to Google Docs and Sheets via a built-in skill or MCP? (Yes → Google Sheet tracker; No → XLSX tracker)
What email account are we working with?
Do you already have Gmail labels set up, even informally?

Outputs

Bible.md in project Drive folder (Markdown — always, regardless of environment choice)
Tracker.xlsx or Google Sheet (depending on environment decision)
_status.md checkpoint file
Summary of label inventory and initial taxonomy draft

1.9 KiB Raw Blame History