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.
npx agentmods add skills/tanstack/table/fuzzy-rankingnpx skills add TanStack/table --skill fuzzy-rankinggit clone --depth 1 https://github.com/TanStack/tableWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00062 | $0.00813 |
| Opus 5 | $0.00031 | $0.00407 |
| Sonnet 5 | $0.00012 | $0.00163 |
| Haiku 4.5 | $0.00006 | $0.00081 |
Grade A, and why
fuzzy-ranking 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Setup
import { compareItems, rankItem, rankings } from '@tanstack/match-sorter-utils'
const query = 'tan'
const values = ['table', 'tanner', 'router']
export const ranked = values
.map((value) => ({
value,
info: rankItem(value, query, { threshold: rankings.MATCHES }),
}))
.filter((entry) => entry.info.passed)
.sort((a, b) => compareItems(a.info, b.info))
Core Patterns
Separate ranking, filtering, and sorting
Call rankItem once, filter on info.passed, retain the RankingInfo, then order matching results with compareItems.
Rank object fields through accessors
type Person = { name: string; email: string }
const person: Person = { name: 'Ada Lovelace', email: '[email protected]' }
const info = rankItem(person, 'lov', {
accessors: [(item) => item.name, (item) => item.email],
})
Accessor options can set a per-accessor threshold plus minRanking/maxRanking bounds.
Store ranking as Table filter metadata
In a Table filterFn, call addMeta?.({ itemRank }). Register the corresponding
meta shape with filterMeta: metaHelper<{ itemRank: RankingInfo }>(). A related
sortFn reads row.columnFiltersMeta[columnId]?.itemRank and uses compareItems,
falling back to an ordinary comparator for ties or absent metadata. For the
primary Table composition, load @tanstack/table-core#global-filtering and
register the fuzzy filter under filterFns for globalFilterFn: 'fuzzy'.
Common Mistakes
HIGH Numeric rank used as pass flag
Wrong: if (rankItem(value, query).rank) include(value).
Correct: test rankItem(value, query).passed.
Ranks below the configured threshold can still be nonzero; passed records the threshold decision.
Source: TanStack/table:packages/match-sorter-utils/src/index.ts
HIGH Ranking recomputed during sorting
Wrong: call rankItem for both rows on every comparator invocation.
Correct: retain RankingInfo during filtering and pass the stored values to compareItems.
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.
- yesterday First seen · 95 lines · 62 tokens per session scan A 1fd9ab0346ba
fuzzy-ranking is a skill published in the GitHub repository TanStack/table (28,393 stars, last pushed 2d ago), licensed MIT. It adds 62 tokens to every session and 813 once invoked, about $0.0003 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.
Other skills, from other repositories
grid-lite-to-igr-grid-migration
Step-by-step migration guide from Grid Lite (IgrGridLite) to the premium Ignite UI for React Data Grid (IgrGrid), covering every import, registration, component name, property, event, template, sorting, filtering, and toolbar API change.
igniteui-react-customize-theme
This skill customizes Ignite UI for React component styling using CSS custom properties and the full theming system and should be used when applying brand colors, dark mode, component-level overrides, or scoped themes in a React application.
webgl-expert
Expert guide for WebGL API development including 3D graphics, shaders (GLSL), rendering pipeline, textures, buffers, performance optimization, and canvas rendering. Use when working with WebGL, 3D graphics, canvas rendering, shaders, GPU programming, or when user mentions WebGL, OpenGL ES, GLSL, vertex shaders…
lenis-react
Integrate Lenis smooth scroll into React and Next.js projects. Use when the user wants smooth scrolling, scroll-linked animations, parallax effects, GSAP ScrollTrigger sync, Framer Motion sync, programmatic scroll-to navigation, or scroll event listening in a React app. Covers ReactLenis provider setup, useLenis hook…
gsap-tackl
Implements and orchestrates GSAP in Tackl — useAnimation, useGSAP, ScrollTrigger, Lenis sync, page transitions, and reduced-motion fallbacks. Use when the user mentions GSAP, tweens, timelines, ScrollTrigger, scrub, matchMedia, Lenis, smooth scroll, page transitions, motion design, or enter/exit animations.
tackl-performance
Optimizes Tackl apps for Web Vitals, reduced motion, and scroll performance. Use when the user mentions performance, LCP, CLS, prefers-reduced-motion, lazy loading, scroll jank, or PerformanceContext.