chini-train-train-0372-dp5-personal

Inbox-Zero Workflow With Vacation Backlog

brutal personal problem: inbox-zero workflow with vacation backlog

Source: chini-train synth generator v0.1

Prompt

Design a system for: inbox-zero workflow with vacation backlog (domain: personal systems / habits).

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.

Cascading failure

cascade

An initial fault propagates through dependent components.

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.

Adversarial burst

adversarial

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

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-train-0372-dp5-personal \
  --provider openrouter --model google/gemini-2.0-flash-001 \
  --as alice

Or inspect the prompt first:

chini-bench prompt chini-train-train-0372-dp5-personal

Providers: openai · anthropic · google · openrouter · ollama

Leaderboard

Rank	Submitter	Model	Score	Stability	Delivery	Design	Pass
#1	rl_v07_full_a10	rl_policy custom single-shot	87	59.0	87.0	100.0	✗
#2	rl_v07_pilot_a10b	rl_policy custom single-shot	84	64.0	64.0	100.0	✗
#3	rl_v07_pilot_a10b	rl_policy custom single-shot	83	62.0	63.0	100.0	✗
#4	rl_v07_full_a10	rl_policy custom single-shot	83	63.0	64.0	100.0	✗
#5	rl_v07_pilot_a10b	rl_policy custom single-shot	82	60.0	61.0	100.0	✗
#6	rl_v07_full_a10	rl_policy custom single-shot	81	58.0	59.0	100.0	✗
#7	chini-train-03	grok-4.1-fast single-shot	80	48.0	67.0	100.0	✗
#8	rl_v07_full_a10	rl_policy custom single-shot	77	55.0	43.0	100.0	✗
#9	rl_v07_pilot_a10b	rl_policy custom single-shot	76	52.0	40.0	100.0	✗

Per-scenario breakdown of the top run

Scenario	Health	Drop rate	Delivered	Pass
baseline	81.0	1.1%	58	✗
cascade-1	23.0	65.3%	13	✗
outage-2	71.0	1.4%	35	✗
latency-3	80.0	0.9%	70	✗
adversarial-4	41.0	100.0%	204	✗

How is this scored? →