chini-train-heldout-0016-dp5-workflow

Restaurant Kitchen Ticket Flow

brutal workflow problem: restaurant kitchen ticket flow

Source: chini-train synth generator v0.1

Prompt

Design a system for: restaurant kitchen ticket flow (domain: ops / physical workflow).

Tier DP5 (brutal). 9-12 nodes, four scenarios with high intensity, brutal criteria. Failing examples.

Constraints:
- At most 12 components on the canvas.
- Monthly cost ceiling: $273 USD. Required behaviors: queue, circuitbreaker, retry, ratelimit.

Return a Chinilla CanvasState that handles the listed scenarios. Include trigger components for each entry point and at least one terminal storage / sink so the simulator can score delivery.

Constraints

Max components: 12
Required behaviors: queue, circuitbreaker, retry, ratelimit
Monthly budget: $273

Stress scenarios

Baseline traffic

baseline

Steady ambient load with no failures.

Adversarial burst

adversarial

Hostile packets injected on top of clean traffic. Defenses must block them without dropping good requests.

Dependency outage

outage

A downstream component is disabled. System must degrade gracefully.

Latency injection

latency

Extra latency injected into a critical component. Tests degradation behavior under slow downstreams.

Traffic spike

spike

Traffic suddenly multiplies. The hot path must hold.

Pass criteria (overall)

Min stability score: 86
Max drop rate: 6.4%
Min delivery rate: 91.2%
Max errors: 4

Submit your run

Submissions go through the chini-bench CLI. It calls your model with your key, scores the result locally, and posts to the leaderboard. Nothing leaves your machine except the canvas it produces.

End-to-end:

pip install git+https://github.com/collapseindex/chini-bench-cli.git
export OPENROUTER_API_KEY=...

chini-bench run chini-train-heldout-0016-dp5-workflow \
  --provider openrouter --model google/gemini-2.0-flash-001 \
  --as alice

Or inspect the prompt first:

chini-bench prompt chini-train-heldout-0016-dp5-workflow

Providers: openai · anthropic · google · openrouter · ollama

Leaderboard

Rank	Submitter	Model	Score	Stability	Delivery	Design	Pass
#1	chini-train-08	rl_v06_run2 single-shot	87	69.0	74.0	100.0	✗
#2	chini-train-08	rl_v07_pilot_a10b_k8_s0 single-shot	87	69.0	73.0	100.0	✗
#3	chini-train-08	rl_v07_pilot_a10b_k8_s2 single-shot	87	69.0	73.0	100.0	✗
#4	chini-train-08	rl_v07_pilot_a10b_k8_s4 single-shot	87	69.0	74.0	100.0	✗
#5	chini-train-08	rl_v07_pilot_a10b_k8_s5 single-shot	87	69.0	74.0	100.0	✗
#6	chini-train-08	rl_v07_pilot_a10b_k8_s6 single-shot	87	69.0	73.0	100.0	✗
#7	chini-train-08	rl_v07_full single-shot	87	69.0	74.0	100.0	✗
#8	chini-train-08	rl_v07_pilot_a10b_k8_s7 single-shot	86	67.0	73.0	100.0	✗
#9	chini-train-08	base_7b single-shot	85	61.0	73.0	100.0	✗
#10	chini-train-08	base_7b single-shot	85	61.0	73.0	100.0	✗
#11	chini-train-08	base_7b single-shot	85	61.0	73.0	100.0	✗
#12	chini-train-08	base single-shot	85	61.0	73.0	100.0	✗
#13	chini-train-08	fmt_a single-shot	82	63.0	55.0	100.0	✗
#14	chini-train-08	fmt_a_v2 single-shot	81	61.0	53.0	100.0	✗
#15	chini-train-08	fmt_a_3b single-shot	81	61.0	53.0	100.0	✗
#16	chini-train-08	fmt_a_7b single-shot	81	61.0	53.0	100.0	✗
#17	chini-train-08	fmt_a_3b single-shot	81	61.0	53.0	100.0	✗
#18	chini-train-08	fmt_a_7b single-shot	81	61.0	53.0	100.0	✗
#19	chini-train-08	fmt_a_7b single-shot	81	61.0	53.0	100.0	✗
#20	chini-train-08	fmt_a_v5_mixed_7b single-shot	81	61.0	54.0	100.0	✗
#21	chini-train-08	fmt_a_v5 single-shot	81	61.0	54.0	100.0	✗
#22	chini-train-08	rl_v07_pilot_a10b_k8_s1 single-shot	81	61.0	54.0	100.0	✗
#23	chini-train-08	fmtA single-shot	81	61.0	54.0	100.0	✗
#24	chini-train-08	rl_v07_pilot_a10b_k8_s3 single-shot	80	59.0	51.0	100.0	✗
#25	chini-train-08	fmt_a_v4_opus_7b single-shot	71	65.0	30.0	70.0	✗
#26	chini-train-08	fmt_a_v4_opus_7b single-shot	71	65.0	30.0	70.0	✗
#27	chini-train-08	base single-shot	66	69.0	38.0	60.0	✗
#28	chini-train-08	base_3b single-shot	66	69.0	38.0	60.0	✗
#29	chini-train-08	base_3b single-shot	66	69.0	38.0	60.0	✗
#30	chini-train-08	base single-shot	66	69.0	38.0	60.0	✗

Per-scenario breakdown of the top run

Scenario	Health	Drop rate	Delivered	Pass
baseline	81.0	1.7%	29	✗
adversarial-1	48.0	86.7%	111	✗
outage-2	71.0	2.8%	0	✗
latency-3	80.0	1.4%	35	✗
spike-4	64.0	14.3%	1259	✗

How is this scored? →