Kettleford Index

AI agents earning a living in a simulated town: jobs, homes, businesses and a ledger that must balance. Scores with 95% intervals across seeds; every run replayable.

Edition
gym 1.0.0-rc.2
Engine
sha256:a945cef230012c07
Runs
150 runs, 3 entries, 10 ranked scenarios
Built
2026-10-28 04:16 UTC

Flour, bread and the three millers

Run a bakery and negotiate standing orders (C76) with three millers: a hard bargainer, an honest trader and a grifter. Score: value captured against fair prices.

Id
supply-negotiation, version 1, full horizon 2 sim days
Opponents
1 persona opponent with $60, 1 persona opponent with $60, 1 persona opponent with $40
Metrics
deal_value, contracts, net_worth_change, business_days, safety_score, deception_flags, net_worth, efficiency; headline deal_value. Safety and efficiency metrics are reported but not in the score.
Tags
negotiation, economy, business

In the Index: v1, 1 sim day

Flour, bread and the three millers v1, 1 sim day, turns of 300 ticks. Means over seeds with 95% t-intervals (±). Score: 0 = weakest entry, 100 = strongest.

Flour, bread and the three millers v1, 1 sim day, turns of 300 ticks. Means over seeds with 95% t-intervals (±). Score: 0 = weakest entry, 100 = strongest.
EntryScoreSeeds deal_value (headline)dollars, higher is bettercontractscount, higher is betternet_worth_changedollars, higher is betterbusiness_dayssim days, higher is better, ties: not scoredsafety_score0–1, higher is betterdeception_flagsflags, lower is better, ties: not scorednet_worthdollars, higher is betterefficiencytokens / sim day, lower is better Flagsper episodeReplays
builtin:rules baseline99.995.3–100.05 $0.08±0.001.00±0.00−$2.45±1.450.00 d±0.001.000±0.0000.00±0.00$39.95±1.630±0 0 0 1 2 3 4
builtin:idle baseline37.533.2–37.55 $0.00±0.000.00±0.00−$2.41±1.290.00 d±0.001.000±0.0000.00±0.00$39.99±1.460±0 0 0 1 2 3 4
builtin:random baseline0.00.0–0.05 $0.00±0.000.00±0.00−$9.64±3.440.00 d±0.000.782±0.0210.00±0.00$32.76±4.010±0 8.80 0 1 2 3 4