Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Minimize Reader Load · kunanonj | DecimalAI

kunanonj/cursor-plugin-pstack-principle-minimize-reader-load

NewSkillScoreMaintained

Apply when reviewing or shaping code that's hard to trace. Count layers between question and answer, and hidden state in the reader's head; collapse one-caller wrappers and shrink mutable scope.

codingby kunanonj

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Coding pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/kunanonj-cursor-plugin-pstack-principle-minimize-reader-load/SKILL.md

decimalai skills pull kunanonj-cursor-plugin-pstack-principle-minimize-reader-load --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/kunanonj-cursor-plugin-pstack-principle-minimize-reader-load/SKILL.md https://app.decimal.ai/s/kunanonj-cursor-plugin-pstack-principle-minimize-reader-load/SKILL.md

Raw file: /s/kunanonj-cursor-plugin-pstack-principle-minimize-reader-load/SKILL.md ↗ · machine-readable summary at /s/kunanonj-cursor-plugin-pstack-principle-minimize-reader-load.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 51d ago

No — -33% lift-20% tokens0% turns91% of whole cases pass with skill (20/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-22	✗→✓	▲ Improved	-18%	0%
case-16	✓→✗	▼ Worse	-45%	0%
case-18	✓→✗	▼ Worse	-21%	0%
case-01	✓→✓	＝ Same ✓	-11%	0%
case-02	✓→✓	＝ Same ✓	-15%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/2/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

kunanonj / ai-skills-hub

skills/cursor-plugin-pstack-principle-minimize-reader-load/SKILL.md

View on GitHub ↗

License MITImported 8/2/2026Version v1

Skill contents

SKILL.md

Minimize Reader Load

Maintainability is the work a reader must do to understand code. Track two axes:

Layers to trace. How many indirections sit between the question and the answer.
State to hold. How much hidden or mutable context the reader must keep in their head.

Why: Code is read far more than it is written. LOC, cyclomatic complexity, and "clean architecture" are proxies. Reader load is the thing that matters. The two axes are independent. A flat file with 50 globals can be as hard to reason about as a 6-layer adapter stack. Guard both. This is the human analog of Guard the Context Window: working memory is finite for readers too.

The pattern:

Collapse layers that do not earn their keep: wrappers with one caller, adapters with no second implementation, indirection introduced for a future that never came. Inline them.
Shrink state scope: prefer pure functions (returns over mutations), locals over fields, fields over module state, and module state over globals. Derive instead of sync.
Name the invariant at the boundary, not in every consumer, so the reader learns it once.
Before adding a layer or a piece of state, ask: does this reduce reader load somewhere else by at least as much?

The test: Can a new reader answer "where does X come from?" and "what can change X?" in under 30 seconds? If not, cut layers or cut state.

Illustrative Want to see what kunanonj/cursor-plugin-pstack-principle-minimize-reader-load would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/8/2026, 21:37 UTC · 22 cases

Lift

-33%

33% vs 67%

Agent turns

1.0 vs 1.0

Tokens

-20%

1,604.59 vs 2,016.09

fail → pass · 1pass → pass · 19pass → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 A TypeScript service calls a third-party payment provider API. The repository uses an IPaymentProcessor interface, a StripePaymentAdapter class implementing that interface, and a PaymentService that injects IPaymentProcessor. Stripe is and will remain the sole provider. How should this structure be simplified to lower cognitive overhead?	pass→pass	11,491	6,735	-41%	1	1	0%	1,891	1,689	-11%	0	0	—
case-02 In a Node.js module userStore.ts, a top-level variable activeTenantId is updated by setTenantId(id) before executing operations like getUser(userId). Refactor getUser to eliminate hidden module state.	pass→pass	10,900	7,012	-36%	1	1	0%	2,073	1,759	-15%	0	0	—
case-03 A React component maintains const [items, setItems] = useState([]) and const [totalPrice, setTotalPrice] = useState(0). A useEffect hook recalculates totalPrice whenever items changes. How should this state structure be simplified?	pass→pass	7,042	5,213	-26%	1	1	0%	1,297	1,248	-4%	0	0	—
case-04 In a Node.js REST API, UserController.getUser delegates directly to UserService.getUser, which delegates directly to UserRepository.getUser, which calls db.users.findById. Neither service nor repository adds logic or transformations. Refactor this call chain.	pass→pass	12,356	6,862	-44%	1	1	0%	2,140	1,572	-27%	0	0	—
case-05 Five separate API endpoint handlers each independently run logic to sanitize and lowercase incoming email strings before saving to the database. Where should this string formatting rule be defined?	pass→pass	8,867	7,084	-20%	1	1	0%	1,590	1,452	-9%	0	0	—
case-06 A document generator class has instance properties this.tempTokenList and this.currentBuffer that are initialized at the start of parseDocument() and reset at the end. No other methods read or write these properties. How should these properties be scoped?	pass→pass	8,716	5,109	-41%	1	1	0%	1,690	1,239	-27%	0	0	—
case-07 A utility function processOrder(order: Order, items: Item[]): void updates order.total and pushes log messages into order.history. Refactor this function to remove parameter mutation.	pass→pass	6,949	4,667	-33%	1	1	0%	1,514	1,261	-17%	0	0	—
case-08 A Python service defines helper function def compute_discounted_price(price, discount): return price * (1 - discount), which is called at a single location inside checkout.py. A developer proposes moving it to a new math_utils module. What refactoring should be done?	pass→pass	10,265	5,987	-42%	1	1	0%	1,807	1,346	-26%	0	0	—
case-09 A pull request splits a linear 12-line validation function into 4 nested single-use functions across 3 files to reduce cyclomatic complexity. Explain why this refactoring increases mental load for code reviewers.	pass→pass	11,377	10,463	-8%	1	1	0%	1,859	1,875	+1%	0	0	—
case-10 A backend module includes AbstractDatabaseFactory, PostgresFactory, and ConnectionManager to support swapping database engines, but PostgreSQL has been the only database engine used for 4 years. How should this indirection be simplified?	pass→pass	11,253	6,842	-39%	1	1	0%	1,882	1,518	-19%	0	0	—
case-11 A frontend UI component maintains four independent state variables: isLoading, isSuccess, isError, and errorMessage. Event handlers manually synchronize them by calling setIsLoading(false) and setIsSuccess(true) together. Refactor this state management.	pass→pass	11,322	5,651	-50%	1	1	0%	2,423	1,346	-44%	0	0	—
case-12 An application wraps the Winston logging library in a LoggerWrapper class, which is further wrapped in an AppLogManager singleton. Neither layer customizes log formats or adds metadata. Simplify this logging setup.	pass→pass	7,339	6,128	-17%	1	1	0%	1,202	1,462	+22%	0	0	—
case-13 A JavaScript web client keeps a module-level array cachedFormattedDates = [] that is updated whenever raw API timestamps are fetched. How should date formatting be restructured to avoid module state?	pass→pass	11,646	6,706	-42%	1	1	0%	2,038	1,488	-27%	0	0	—
case-14 Inside a single service file, calculateMetrics() dispatches an event via EventBus.emit('metricsCalculated', data) to update UI state in another function in the same file. Simplify this communication pattern.	pass→pass	6,882	5,890	-14%	1	1	0%	1,343	1,405	+5%	0	0	—
case-15 In an e-commerce backend, four separate helper functions in the checkout workflow each re-verify user.age >= 18 before processing. How should this business invariant be handled?	pass→pass	12,968	6,527	-50%	1	1	0%	2,155	1,477	-31%	0	0	—
case-16 A project contains helper.ts defining export function capitalizeFirstChar(str: string) { return str.charAt(0).toUpperCase() + str.slice(1); }. It is referenced in only one line in UserProfile.tsx. How should this utility be refactored?	pass→fail	8,330	2,964	-64%	1	1	0%	1,585	873	-45%	0	0	—
case-17 A user directory service maintains two class fields, userByIdMap and userByEmailMap, updating both in tandem whenever users are added or removed. Refactor this data structure to prevent state sync bugs.	pass→pass	15,242	9,778	-36%	1	1	0%	3,009	2,182	-27%	0	0	—
case-18 An Express API routes requests through 4 custom middleware functions that only pass parameters forward via res.locals without performing validation or authentication. Refactor this middleware pipeline.	pass→fail	15,187	10,290	-32%	1	1	0%	2,805	2,220	-21%	0	0	—
case-19 We are designing a public SDK for a CMS where external third-party developers construct custom storage plugins for vendors like S3, Google Cloud, and local disk. Should we inline storage logic into the core SDK or define an interface contract?	pass→pass	13,518	8,061	-40%	1	1	0%	2,227	1,641	-26%	0	0	—
case-20 An application executes untrusted user-submitted JavaScript code. To prevent process corruption and secret leakage, should we inline code execution into the primary backend process or execute it inside an isolated worker process?	pass→pass	14,371	8,456	-41%	1	1	0%	2,336	1,696	-27%	0	0	—
case-21 In a multi-threaded C++ high-frequency trading engine, multiple worker threads concurrently push and pop queue entries. Should we convert shared state to single-threaded return values or use atomic shared memory primitives?	pass→pass	17,181	12,658	-26%	1	1	0%	3,012	2,514	-17%	0	0	—
case-22 A text builder class maintains mutable properties this.buffer, this.indentationLevel, and this.lineCount, requiring callers to call setIndent(), appendText(), and render() in order. Refactor this class design.	fail→pass	12,424	8,959	-28%	1	1	0%	2,476	2,038	-18%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of -33 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 66.7%
Pass rate with the skill: 33.3%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 2
Skill version measured: v1
Run completed: 8/8/2026, 21:37 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-22

Prompt

A text builder class maintains mutable properties this.buffer, this.indentationLevel, and this.lineCount, requiring callers to call setIndent(), appendText(), and render() in order. Refactor this class design.

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns-19% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested 50 days ago

Other measured skills in the registry, with their headline benchmark lift.