>_ build log #001 / audio summary

Listen — how the session system was built, in 2 minutes

0:00 / 0:00

narration: matt browning · synthesised via elevenlabs

>_ gtmstack.ai / system map / v0.2

# session pipeline

How a stack session actually runs. The voice trigger fires session-starter, which silently fetches gtmstack.ai + 6 repo sources, cross-references your Google Calendar 1-2-1s, replays open actions and parked items, surfaces ADR conflicts and 7-day gaps, then categorises every input by persona — Marketing / Technical / Business — by priority and by repo impact. Output is a session log, generated diagrams, and an email of the decisions.

>_ 26 nodes 17 live 2 in progress 7 plannedv0.2 — built live
>_ 01 / input2/3 live
>_ 02 / skill activation3/3 live
>_ 03 / processing2/4 live
>_ 04 / output1/1 live
>_ 05 / session close3/3 live
>_ 06 / airtable data layer0/4 live
gtmstack — source of truth
>_ 07 / github4/4 live
Marketing-Matt/gtmstack
>_ 08 / content engine + gtmstack.ai2/4 live

# status

  • live — solid border
  • planned — dashed
  • in progress — solid grey

# layer tints

  • 02 / skill activation
  • 03 / processing
  • 04 / output

# interactions

  • click node → details
  • filter chips → dim others
  • esc → close panel
*>_ gtmstack.ai — Unfiltered AI marketing. Built live.*