Why Not a Monorepo Tool

What this means for you

The honest competitor is not a build tool — it is the tab in your wiki where you keep “how this project actually works” and the message to a teammate that says “let me re-explain the architecture again”. Build tools cache test runs. Petrova caches project context for the next agent session.

The pitch

A monorepo tool decides what to rebuild when a file changes. Petrova decides what an agent should do next when a session opens. Different layer, different problem — and you can run both.

Who it’s for

The team that already has Nx or Turborepo running their builds, but still spends the first ten minutes of every AI session re-explaining what the project is.

Proof points

• Build tools read your dependency graph; petrova reads your spec docs and your phase state
• Petrova ships zero CI logic — your existing pipelines keep working unchanged
• The deliverable is three plain markdown / XML files an agent reads, not a binary or a daemon
• Works alongside, not instead of: an Nx workspace can register itself with petrova and adopt the playbook in one bootstrap session

quadrantChart
  title Petrova vs. adjacent tools
  x-axis "Caches build artifacts" --> "Caches project context"
  y-axis "For humans only" --> "For AI agents too"
  quadrant-1 "Petrova"
  quadrant-2 "Wiki / handbook"
  quadrant-3 "Hand-rolled bash"
  quadrant-4 "Monorepo build tools"
  "Nx": [0.15, 0.25]
  "Turborepo": [0.2, 0.2]
  "Notion handbook": [0.7, 0.15]
  "Hand-rolled scripts": [0.35, 0.3]
  "Petrova": [0.85, 0.9]