A fintech API platform at Walk pace with 8 completed items of history. Three specialists with dense cross-team pathways, tuned knowledge distribution, multi-floor governance (compliance + behavioral), and overrides on scrum-master and CKO reflecting lessons learned. This is what the framework looks like after it’s had time to learn.
setup.shsetup.sh — compiles both floors during setup05-operational-maturity/
├── .claude/
│ ├── agents/
│ │ ├── frontend-specialist.md # Fintech frontend (extends harness)
│ │ ├── backend-specialist.md # Fintech backend (extends harness)
│ │ └── e2e-test-engineer.md # Full-stack E2E testing (extends harness)
│ └── overrides/
│ ├── scrum-master.md # Retro every 3, pace monitoring
│ └── cko.md # Walk-pace knowledge distribution
├── floors/
│ ├── compliance.md # 5 rules + 3 enforcement blocks (fintech-oriented)
│ └── behavioral.md # 2 rules + 1 enforcement block (process quality)
├── fleet-config.json # Walk pace, 3 specialists, multi-floor, dense pathways
├── ops/
│ ├── checks/
│ │ └── verify-transactions.sh # Custom enforcement script (transaction validation)
│ └── deploy.sh # Deploy stub with idempotency/E2E checks
└── setup.sh # Seeds metrics history, compiles floors
ops/test-example.sh 05-operational-maturity
# cd into the worktree, open Claude Code
ops/dora.sh — See a dashboard with real history (seeded by setup.sh)ops/dora.sh --sm — See the SM’s pace recommendation based on metricsops/dora.sh --flow — See handoff quality across specialist boundaries/po — Notice the fleet is at Walk pace, not Crawlops/pathways.sh — Compare declared vs actual agent communicationYou should be comfortable with all previous examples. This one shows the operational destination — what earlier examples are building toward.