Scoring

faf score reads your project.faf and gives it a 0–100% AI-readiness score. The score is deterministic: the same file gets the same number, every time, in every tool that runs the engine.

faf score           # score + every slot
faf score --json    # the same, for scripts

Always-33

Every .faf is scored against the same 33 slots:

Slots	What they hold
project (3)	name, goal, main language
human context (6)	who, what, why, where, when, how
frontend (4) · backend (5) · universal (3)	the stack
enterprise (12)	infra, app and ops

Score = filled ÷ active. A slot is active unless it's marked slotignored.

Your app-type decides what counts

project.type is your app-type. faf init, faf auto and faf git detect it and write slotignored into every slot it doesn't use — a CLI counts 12 slots, a full-stack app 21. Everything else stays active: filled, or empty.

The slots you don't use stay visible. faf score lists all 33:

✪ TROPHY 100% 12/12 slots — project.faf
  ● project.name
  …
  ● stack.cicd
  — stack.monorepo_tool: slotignored
  …
  — monorepo.remote_cache: slotignored

What counts as empty

A missing slot, an empty value, or a placeholder word — None, N/A, null, unknown, TBD, TODO, not applicable — counts as empty. Only slotignored takes a slot out of the score, and faf writes it from your app-type.

A file without the enterprise markers

If a .faf doesn't mark the 12 enterprise slots, they count as empty: 21 filled is 21 ÷ 33 = 64%. Run faf auto — it writes the markers, and the score is what your file actually says.

Tiers

Score	Tier
100%	✪ Trophy
99%	★ Gold
95%	◆ Silver
85%	◇ Bronze
70%	● Green
55%	● Yellow
1%	○ Red
0%	♡ White

Trophy is the target: at 100%, AI never has to guess.