Skip to the content.

fak maturity scorecard — lifecycle, not just completeness

maturity_debt (ladder-skips): 0; maturity index 50/100 (F) over 583 declared capabilities (102 carry the benchmarked badge).

maturity: 583 capabilities, index 50/100 (F); no ladder-skips (every capability is at most as mature as its evidence); 2 proposed · 1 prototyped · 575 tested · 5 dogfooded · 0 default

A v1 prototype can be legitimately complete and still not be tested, dogfooded, benchmarked, or the default. This scorecard makes that lifecycle visible: it places every declared capability (one per internal/<leaf> lane in dos.toml [lanes.trees]) on a closed ladder, and for each one names the next step that would mature it. Every rung is gated by evidence the author did not write — code on disk, a *_test.go, an edge in the running binary’s transitive import graph (fak itself runs it), a documented verb — so the only way up the ladder is to change the real tree.

Immaturity is not a defect. A capability honestly at prototyped is a complete v1 that simply has not been matured yet — expected, and never counted against anyone. The one defect this refuses is a ladder-skip: a capability that looks more mature than its evidence — concretely, one fak relies on (dogfooded, a default surface, or benchmarked) yet leaves untested. That is the maturity sibling of the product scorecard’s verdict-overclaim and the readiness ladder’s READINESS_OVERCLAIM (#582 / grammar G1).

The lifecycle ladder

# Rung Reached when (evidence the author did not write)
0 proposed a declared capability with no code on disk yet
1 prototyped a non-test .go file exists in the leaf — a complete v1
2 tested the leaf carries a *_test.go (the QA rung)
3 dogfooded the leaf is on the running binary’s transitive import graph — fak itself runs it
4 default a passing runtime proof declares the capability active without an opt-in action (default_on=true)
· benchmarked (badge) a func Benchmark* in the leaf or a BENCHMARK-AUTHORITY.md row — the natural step after default

Distribution

Rung Capabilities
proposed 2
prototyped 1
tested 575
dogfooded 5
default 0
benchmarked (badge) 102

Next work — the agentic-culture backlog

Each gap is a concrete, checkable next work item. fak maturity next is the queue an agent (or the issue-dispatch loop) pulls from to advance the fleet one rung at a time. Ladder-skips first (they are the real debt), then the least-mature capabilities (the most leverage).

  From → gap Next work item Witness
  proposed → prototyped prototype framebus: land a v1 in internal/framebus a non-test .go file exists under internal/framebus
  proposed → prototyped prototype kimik3page: land a v1 in internal/kimik3page a non-test .go file exists under internal/kimik3page
  prototyped → tested test xprobe: add unit tests covering internal/xprobe a *_test.go in internal/xprobe (go test ./internal/xprobe/… passes)
  tested → dogfooded dogfood a2achan: exercise a real runtime path and record its passing command a passing runtime command recorded for a2achan in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood abi: exercise a real runtime path and record its passing command a passing runtime command recorded for abi in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood accountobs: exercise a real runtime path and record its passing command a passing runtime command recorded for accountobs in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood accountprobe: exercise a real runtime path and record its passing command a passing runtime command recorded for accountprobe in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood accounts: exercise a real runtime path and record its passing command a passing runtime command recorded for accounts in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood adjudicator: exercise a real runtime path and record its passing command a passing runtime command recorded for adjudicator in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood advmodel: exercise a real runtime path and record its passing command a passing runtime command recorded for advmodel in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood affectedtests: exercise a real runtime path and record its passing command a passing runtime command recorded for affectedtests in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agent: exercise a real runtime path and record its passing command a passing runtime command recorded for agent in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agentdemo: exercise a real runtime path and record its passing command a passing runtime command recorded for agentdemo in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agentdojo: exercise a real runtime path and record its passing command a passing runtime command recorded for agentdojo in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agenticbench: exercise a real runtime path and record its passing command a passing runtime command recorded for agenticbench in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agentreadinessscore: exercise a real runtime path and record its passing command a passing runtime command recorded for agentreadinessscore in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agentsindex: exercise a real runtime path and record its passing command a passing runtime command recorded for agentsindex in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agenttest: exercise a real runtime path and record its passing command a passing runtime command recorded for agenttest in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood agenttopo: exercise a real runtime path and record its passing command a passing runtime command recorded for agenttopo in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood ailuminate: exercise a real runtime path and record its passing command a passing runtime command recorded for ailuminate in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood amdgpu: exercise a real runtime path and record its passing command a passing runtime command recorded for amdgpu in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood answershape: exercise a real runtime path and record its passing command a passing runtime command recorded for answershape in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood antipattern: exercise a real runtime path and record its passing command a passing runtime command recorded for antipattern in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood apihostprobe: exercise a real runtime path and record its passing command a passing runtime command recorded for apihostprobe in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood appversion: exercise a real runtime path and record its passing command a passing runtime command recorded for appversion in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood architest: exercise a real runtime path and record its passing command a passing runtime command recorded for architest in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood archreport: exercise a real runtime path and record its passing command a passing runtime command recorded for archreport in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood assumecheck: exercise a real runtime path and record its passing command a passing runtime command recorded for assumecheck in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood astquery: exercise a real runtime path and record its passing command a passing runtime command recorded for astquery in internal/maturity/runtime-proofs.json
  tested → dogfooded dogfood atif: exercise a real runtime path and record its passing command a passing runtime command recorded for atif in internal/maturity/runtime-proofs.json
    … and 553 more (run fak maturity next)  

Run it

go run ./cmd/fak maturity              # the lifecycle scorecard
go run ./cmd/fak maturity next         # the next-work backlog (ladder-skips first)
go run ./cmd/fak maturity route        # plan deduped public GitHub issues for the top public rows
go run ./cmd/fak maturity route --live # create/update public-routeable issues so dispatch can drain them
go run ./cmd/fak maturity --markdown    # regenerate this doc
go run ./cmd/fak maturity --json        # machine payload (control-pane / dispatch loop)
go test ./internal/maturity/...        # prove the ladder + skip detection + next-work fold

maturity route keeps private-boundary lanes visible in maturity next, but reports them as skipped instead of filing public issues. The default-on maturity-continuity cycle runs the same full-portfolio invariant weekly.

Next: advance the fleet one rung: fak maturity next lists 583 next work item(s); the least-mature capability is the most leverage