The Post-AI PM Playbook
  1. Preface · It was always about product sense
  2. 01 · You can only sell what you understand
  3. 02 · The product operating model as your shared context layer
  4. 03 · Composition, not headcount
  5. 04 · Guardrails to reduce assumptions and hallucinations
  6. 05 · Two moments for the humans in the loop
  7. 06 · Run it as loops, not projects
  8. How to start

Maskin

Maskin implements this playbook.

Maskin ships this playbook as executing infrastructure.

Agent-native product management is the discipline the chapters describe. Maskin is the agentic workspace where that discipline runs: humans keep the judgment, agents do the closing, one shared typed object graph makes team memory compound with every bet.

The loop, running as product

  • Signal. Insights land in the workspace as typed objects, from users, agents, and the market. Nothing gets buried in chat or lost between tools.
  • Bet. A shaped bet with a close condition frames the work. The human gate holds taste on the frame; agents pick up the rest.
  • Shipped outcome. The loop closes on a validated result. Team memory compounds. The next bet inherits the context, not just the code.

Open-source, MCP-native, self-hostable, BYO model. Apache 2.0.

See Maskin →