loop-cost

Estimate daily token spend for agent loops patterns by cadence and readiness level (L1–L3).

Uses cost metadata from patterns/registry.yaml.

Install & Run

npx @kevinzhangnothing/loop-cost --pattern ci-sweeper --cadence 15m --level L2
npx @kevinzhangnothing/loop-cost --pattern daily-triage --level L1 --json
npx @kevinzhangnothing/loop-cost --list

From this repo:

cd tools/loop-cost
npm install
npm test

Options

| Flag | Description | |------|-------------| | --pattern | Pattern id (see --list) | | --cadence | Override cadence (e.g. 15m, 1d) | | --level | L1, L2, or L3 (default L1) | | --conservative | Use slower cadence from ranges | | --json | Machine-readable output |

Scenarios

Each estimate includes:

• Early-exit / no-op — empty watchlist, minimal tokens
• Full triage — every run does a full scan
• Action every run — implementer + verifier every time (worst case)
• Realistic blend — level-based mix (documented in output)

Pair with loop-budget.md (scaffolded by loop-init) and loop-audit cost observability checks.

Feed the circuit breaker

--json output resolves a realistic per-run token cap for loop-context's breaker, so --token-budget reflects real pattern cost instead of a hand-typed guess (the loop-guard skill wires this):

BUDGET=$(npx @kevinzhangnothing/loop-cost --pattern ci-sweeper --level L2 --json \
  | node -e 'let d="";process.stdin.on("data",c=>d+=c).on("end",()=>process.stdout.write(String(JSON.parse(d).scenarios.realistic.tokensPerRun)))')
npx @kevinzhangnothing/loop-context --check --ledger loop-ledger.json --token-budget "$BUDGET"

The tools stay independent — no runtime dependency, just shell wiring.

See docs/operating-loops.md.