opensearch-schema-review
Skill1 starsChanged 56 days agoApache-2.0View on GitHub
Review OpenSearch 3.5 index mappings/schemas. Check the dynamic policy, text/keyword, multi-field, index/doc_values, _source/derived source, nested cost, knn_vector fields (dimension/ space_type/engine/method/mode), and the Korean nori analyzer. Use when designing/modifying a mapping or reviewing an index schema. Grounds findings in a KB based on the official OpenSearch docs.
Files
OpenSearch Schema/Mapping Review (Schema Review)
The grounding principles are in reference/principles.md. For facts, prefer reference/kb/ (the official docs).
Review order (proceed in this order)
- What index does this index? — Identify the domain, language (Korean?), time-series nature, and scale (N).
- Mapping in general — Check the review hooks in
reference/kb/field-types.md·reference/kb/dynamic-mapping-risks.md. - Analyzers — Check
reference/kb/analyzers-tokenizers.md(search vs index analyzer, normalizer, nori). If the nori grep returns 0 hits, state "not applicable". - Vector fields — Check
reference/kb/knn-vector-field.md(dimension/space_type/engine/mode consistency, immutability). - Templates —
reference/kb/index-templates.md. Along with the severity, provide a fix (mapping JSON diff) + a zero-downtime reindex/alias strategy.
Quick checklist
- Is
dynamicset tostrict/falseon production indexes (to prevent mapping explosion)? - Is sorting/aggregation/exact on
keywordand full-text search ontext? Are unused fieldsindex:false/doc_values:false? - Does the
knn_vectordimensionmatch the model output, and is thespace_typeconsistent with normalization? - Is the
enginefaiss/lucene (migrate if nmslib)? Are the core fields explicit in the mapping (to prevent dynamic inference)? - Are large vectors duplicated in
_source→ derived source (beware that nested/copy_to are unsupported)? - (Korean) Is nori configured with
decompound_mode, a user dictionary, parts of speech, and a search-timesynonym_graph?
KB (read and cite first)
From reference/kb/INDEX.md, read the following, check against the "review hooks", and cite via the source URL:
reference/kb/field-types.md·reference/kb/analyzers-tokenizers.md·reference/kb/knn-vector-field.md·reference/kb/index-templates.md·reference/kb/dynamic-mapping-risks.md
Deliverable
For each finding: problem → violated principle (principles.md §)/KB source → mapping diff → verification (_mapping/_analyze). Because mappings are immutable, accompany any change with a zero-downtime reindex/alias strategy.
In this repository65 other files it gives its agents.
- agent-evalSkill
- architecture-review-workflowSkill
- backend-code-review-workflowSkill
- code-authoring-routerSkill
- code-diagram-extractSkill
- codex-agent-runtimeSkill
- design-pattern-applySkill
- diagram-stylerSkill
- document-designerSkill
- e2e-evidence-reportSkill
- e2e-report-rendererSkill
- e2e-runnerSkill
- e2e-scenario-authorSkill
- e2e-test-workflowSkill
- entitySkill
- enumSkill
- explore-dead-codeSkill
- flywaySkill
- git-flowSkill
- harness-proposeSkill
- harness-refineSkill
- java-spring-review-workflowSkill
- knowledge-auditSkill
- korean-dev-writerSkill
- kotlin-code-authoringSkill
- kotlin-spring-reviewSkill
- kotlin-spring-review-workflowSkill
- learnSkill
- mapperSkill
- markdown-publisherSkill
- mermaid-diagramsSkill
- mssql-code-authoringSkill
- nl-sql-querySkill
- notion-domain-contextSkill
- oop-design-reviewSkill
- opensearch-code-authoringSkill
- opensearch-optimizationSkill
- opensearch-query-reviewSkill
- opensearch-review-workflowSkill
- pdf-qaSkill
- playwright-e2eSkill
- postgres-code-authoringSkill
- postgres-query-reviewSkill
- postgres-review-workflowSkill
- postgres-risk-checkSkill
- postgres-schema-designSkill
- product-requirementsSkill
- product-ui-designSkill
- refactoring-catalogSkill
- repositorySkill
- requestSkill
- responseSkill
- retroSkill
- schema-doc-extractSkill
- security-scan-workflowSkill
- serviceSkill
- skill-updaterSkill
- start-taskSkill
- test-authoringSkill
- test-quality-refactoring-workflowSkill
- The first 60 shown.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

