Every premium and members-only resource for $9.99/month, as a no-commitment subscription. See the plans

PalateKit

Sign in
Menu

Request a resource

Describe the resource you are missing in one sentence. We will email you when it is published.

0 / 280

A page, a video or an image online.

Unlock PalateKit

Unlock everything

Plan

Paying means accepting our terms of sale.

Secure payment by Stripe. Receipts and invoices sent by email.

Welcome back

Sign in to install, download or save this resource. A free account is enough for free resources.

OpenSearch scoring functions for marketplace ranking v1.0.0

Use this resource to design an OpenSearch ranking for a marketplace with scoring functions and evaluation

New

  • 316 downloads
  • 147 likes
  • 119 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install opensearch-ranking-functions

With no option, the skill goes to ~/.agents/skills/; add --agent claude-code (or codex, cursor…) to target an agent.

First time: connect the CLI with

npx --yes palatekit@latest login

The agent designs or reviews an OpenSearch ranking against 56 rules: hybrid BM25 and kNN retrieval, quality signals, spatial or temporal decay, supply and demand balance, bias correction and evaluation with NDCG or A/B tests. It states the expected gain of each function. You get a measured ranking rather than a score tuned by feel.

Mostly picked by

Updated November 2, 2026

Card for the “OpenSearch scoring functions for marketplace ranking” skill

Details

Works in
Install guides
Generates with
Your agent's model
Licence
MIT

Similar resources

View all

What you get

  • references/: cards on bias correction, click models and Thompson sampling
  • references/: cards on Gaussian and exponential decay and how to combine them

Good fit for

  • Recent listings are buried under older, well-rated ones
  • A function_score was tuned by hand without any measurement

Not for

    Tags

    • opensearch
    • ranking
    • marketplace
    • ndcg
    • scoring

    About

    You say

    “Our rental ranking favours old listings. Propose a time decay and a way to measure it with NDCG.”

    Card for the “OpenSearch scoring functions for marketplace ranking” skill

    Skill published by pproenca under the MIT license, delivered in its original version.

    Show more

    How to use it

    1. Install the skill in your agent's skills folder
    2. Ask your agent to design an OpenSearch ranking for a marketplace with scoring functions and evaluation
    3. Review and adjust the result

    Content preview

    Files

    • AGENTS.md19 KB
    • assets/templates/_template.md1.7 KB
    • LICENSE1.1 KB
    • metadata.json5.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • references/_sections.md4.5 KB
    • references/bias-click-models.md3.3 KB
    • references/bias-counterfactual-eval.md4.5 KB
    • references/bias-interleaved-evaluation.md4.7 KB
    • references/bias-popularity-debiasing.md4 KB
    • references/bias-position-ips.md4.1 KB
    • references/bias-thompson-sampling.md4.4 KB
    • references/decay-exp-freshness.md3.4 KB
    • references/decay-gauss-geo.md3.3 KB
    • references/decay-multi-field-composition.md4 KB
    • references/decay-offset-noise.md2.8 KB
    • references/decay-scale-calibration.md3.2 KB
    • references/div-category-diversity.md4.1 KB
    • references/div-dpp-quality-diversity.md4.3 KB
    • references/div-max-per-host.md3.5 KB
    • references/div-mmr-rerank.md4.3 KB
    • references/div-window-penalty.md4 KB
    • references/eval-ab-sample-size-mde.md5.1 KB
    • references/eval-ablation-attribution.md5.3 KB
    • references/eval-cuped-variance-reduction.md5.4 KB
    • references/eval-graded-judgment-set.md4.3 KB
    • references/eval-ndcg-primary-metric.md4.6 KB
    • references/eval-online-offline-correlation.md4.9 KB
    • references/eval-regression-query-suite.md5.6 KB
    • references/market-cold-start-exploration.md3.9 KB
    • references/market-conversion-weighted-ranking.md3.4 KB
    • references/market-host-quality-signals.md4 KB
    • references/market-inventory-health.md3.7 KB
    • references/market-pareto-multi-objective.md4.2 KB
    • references/market-price-relevance.md4 KB
    • references/market-supply-fairness-lorenz.md4.1 KB
    • references/pers-contextual-features.md3.5 KB
    • references/pers-cross-encoder-rerank.md3.2 KB
    • references/pers-listing-embeddings.md3.7 KB
    • references/pers-multi-modal-embeddings.md3.6 KB
    • references/pers-real-time-session-vector.md3.5 KB
    • references/pers-tower-split-offline-online.md3.9 KB
    • references/pers-type-embeddings-cold-start.md3.5 KB
    • references/qual-bayesian-average.md3.1 KB
    • references/qual-completeness-score.md3.1 KB
    • references/qual-log1p-vs-saturation.md3 KB
    • references/qual-rank-feature-saturation.md3 KB
    • references/qual-rank-feature-sigmoid.md2.9 KB
    • references/qual-wilson-lower-bound.md2.8 KB
    • references/recall-hnsw-vs-ivf.md2.4 KB
    • references/recall-hybrid-rrf.md2.3 KB
    • references/recall-multi-stage.md2.4 KB
    • references/recall-prefilter-knn.md2.3 KB
    • references/recall-query-expansion.md2.8 KB
    • references/recall-two-tower-ebr.md2.5 KB
    • references/rel-avoid-boost-inflation.md2.6 KB
    • references/rel-bm25-k1-b-tuning.md2.7 KB
    • references/rel-bm25f-field-weights.md2.4 KB
    • references/rel-listwise-loss.md2.7 KB
    • references/rel-multi-match-strategy.md2.5 KB
    • references/rel-rescore-over-bool-should.md3.2 KB
    • references/rel-script-score-over-function-score.md3.2 KB
    • SKILL.md15.1 KB

    Origin and licence

    Listed unmodified

    Author
    pproenca
    Source
    github.com/pproenca/dot-skills/tree/cf93c57cac89d6fc3e4194686000411567f5caf3/skills/.experimental/opensearch-function-scoring-algorithms
    Licence
    MIT

    Notice

    « opensearch-function-scoring-algorithms » (pproenca, MIT) livré dans sa version d'origine : fiche en français et en anglais par PalateKit.

    User reviews

    –

    No ratings yet

    Loading reviews…

    Delete your review?

    Your rating and comment will be removed from the page.