cost-benchmark

Cost Benchmark

Runs scripts/bench.mjs against the structural+adversarial corpus and writes per-case + summary results to docs/benchmarks/runs/. This is the verification gate that backs every measurable claim in cost-booster-edit / cost-booster-route.

When to use

• Before publishing a release — verify booster win rate didn't regress.
• After expanding bench/booster-corpus.json — confirm new cases route correctly.
• When auditing a "claimed upstream" tag — flip it to "verified" once the bench supports it.
• On a cost question ("is Sonnet 4.6 cheaper than Opus 4.7 for these tasks?") — re-run with BENCH_ANTHROPIC=1.

Steps

1. Run the bench from v3/ (where agent-booster resolves):

   ( cd v3 && node ../plugins/swarmdo-cost-tracker/scripts/bench.mjs )                  # booster only — free, ~85 ms
   ( cd v3 && BENCH_LLM_BASELINE=1 node ../plugins/swarmdo-cost-tracker/scripts/bench.mjs ) # + Gemini 2.0 Flash (cheap)
   ( cd v3 && BENCH_LLM_BASELINE=1 BENCH_ANTHROPIC=1 \
        node ../plugins/swarmdo-cost-tracker/scripts/bench.mjs )                          # + Sonnet 4.6 + Opus 4.7
   

2. Inspect the markdown summary printed to stdout. The gate metric is winRate (Tier 1 cases). Adversarial cases are tracked separately as escalationRate.

3. Persisted output lands at:

   • docs/benchmarks/runs/latest.json — pointer to the most recent run
   • docs/benchmarks/runs/<ISO-timestamp>.json — historical record

4. Read it back in subsequent skills (e.g. cost-report step 2 reads latest.json for live tier-spend numbers).

Smoke gates

• winRate ≥ 0.80 on Tier 1 cases (smoke step 23). Lower the threshold by editing scripts/smoke.sh.
• escalationRate is reported but ungated — adversarial cases are diagnostic.

Env overrides

Env var	Default	Purpose
BENCH_LLM_BASELINE	unset	=1 runs the OpenAI-compat baseline
BENCH_LLM_MODEL	models/gemini-2.0-flash	Override the OpenAI-compat model
BENCH_LLM_BASE_URL	Gemini OpenAI shim	Override endpoint
BENCH_ANTHROPIC	unset	=1 runs Anthropic baseline (Sonnet 4.6 + Opus 4.7)
BENCH_ANTHROPIC_MODELS	claude-sonnet-4-6,claude-opus-4-7	Comma-separated Claude IDs
BENCH_OUT	timestamped file	Override output path
BENCH_QUIET=1	unset	Suppress markdown summary

API keys auto-pulled from gcloud secrets (GOOGLE_AI_API_KEY, ANTHROPIC_API_KEY); override with BENCH_LLM_API_KEY / BENCH_ANTHROPIC_API_KEY.

Cross-references

ADR-0002 §"Decision 1" / §"Riskiest assumption" · cost-booster-edit/SKILL.md (verification table consumes this skill's output) · cost-report/SKILL.md step 2 (reads runs/latest.json).