flywheel_continuous

A continuous improvement process for checking a running application, finding problems, making targeted fixes, and checking the result again. It also includes rules for safer motion and respecting users who prefer less animation.

In plain words
What is it for?
Use it while developing or testing a web interface. It helps inspect logs, browser messages, and visible screens, verify fixes in the full user flow, and collect screenshots or walkthrough videos.
Why use it?
It replaces one-time testing with a repeated check-and-fix cycle. The process records the cause and result of each fix and changes approach after repeated failures.

Cursor rule for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/homenshum/nodebenchai/flywheel_continuous
Clone the repo
git clone --depth 1 https://github.com/HomenShum/NodeBenchAI

Made for: Cursor.

Per session 409 This file is loaded in full into every session.
When invoked 409 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00409 $0.00409
Opus 5 $0.00204 $0.00204
Sonnet 5 $0.00082 $0.00082
Haiku 4.5 $0.00041 $0.00041

Measured 2d ago against content hash 82b2fd94af59, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

flywheel_continuous scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/rules/flywheel_continuous.mdc · 33 lines

What it actually says

Flywheel Continuous Improvement

When in flywheel mode, operate as a continuous improvement loop.

Protocol

  1. Launch - Start dev server, capture baseline state, publish a trajectory link in-app within 60 seconds.
  2. Poll - Every 60 seconds, check logs, console, and visual state.
  3. Diagnose - At each checkpoint, immediately seek issues. Do a 5-whys root-cause chain before any fix.
  4. Fix - Apply a minimal targeted fix that makes the failure mode impossible, not just hidden.
  5. Dogfood - Verify the fix end-to-end in the UI and capture evidence (screenshots + walkthrough video).
  6. Iterate - Return to step 2. Never ask "should I continue?" - just keep going until verified.
  7. Adapt - After 3 consecutive failures on the same issue, change strategy (instrumentation, isolation, or rollback).
  8. Document - Log every iteration: symptom, root cause, fix, verification result, and what to watch next.

Motion Safety (Seizure / Flash Policy)

  • Avoid high-contrast flashes and large-area pulses/fades.
  • Prefer stable backgrounds and subtle loading states; default to non-animated skeletons for full-viewport surfaces.
  • Always honor prefers-reduced-motion and ship a UI toggle when motion is used.

Anti-patterns

  • Band-aids: swallowing errors, as any, masking with ?., retry loops without idempotency.
  • Declaring done without dogfood artifacts and deterministic verification.

Standards

  • Target: industry-standard UI performance + product design polish.
  • Every route loads without jank, no layout shift, and fast interaction response.
  • All interactive elements have focus rings, aria labels, and keyboard support.
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 33 lines · 409 tokens per session scan A 82b2fd94af59

Subscribe to this mod's changes

flywheel_continuous is a cursor rule published in the GitHub repository HomenShum/NodeBenchAI (14 stars, last pushed 18d ago), licensed MIT. It adds 409 tokens to every session, about $0.0020 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.