convez.ai
SkillDados & AnalyticsIntermediário
celticht32

Couchbase Skills para Claude

celticht32/Couchbase-Skills-for-Claude.ai

29 skills de contexto para dominar Couchbase com Claude em todos os cenários

Ver no GitHub
InicialPluginScore · de 0 a 100

Coleção de 29 skills estruturadas para Claude que cobrem operações com Couchbase, desde integração de aplicações até IA, Kubernetes, sincronização mobile, segurança e analytics. Cada skill encapsula padrões, anti-padrões e árvores de decisão específicas do Couchbase que permitem ao Claude fornecer respostas confiáveis e contextualizadas.

Instalação

$npx convez-ai celticht32/Couchbase-Skills-for-Claude.ai

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
4
Forks
2
Acompanhando
0
Saúde
57%
Último commit
há 4 meses
Versão
—
Issues abertas
1
Contribuidores
1

Casos de uso

  • Otimização e debugging de queries SQL++ em Couchbase
  • Design e modelagem de dados para aplicações NoSQL
  • Deployment, operações Kubernetes e sincronização mobile
  • Integração de Couchbase com aplicações e APIs
  • Hardening de segurança e configuração de clusters

Veredito

Recurso bem-estruturado e maduro para quem trabalha com Couchbase em produção. A modularidade permite carregar apenas o contexto necessário, evitando desperdício de tokens, e a cobertura é abrangente — de queries até DevOps. Requer familiaridade com Couchbase para extrair valor máximo.

README

A tradução para o português deste README ainda está na fila.

Couchbase Claude Skills

Claude skill files for working with Couchbase — covering every major service and deployment pattern from application integration through AI applications, Kubernetes operations, mobile sync, security hardening, and analytics.

29 skills · 124 files · ~18,000 lines

License: MIT


What are Claude Skills?

Skills are structured context files (SKILL.md + optional references/ docs) that Claude loads before responding to requests in a specific domain. They encode patterns, anti-patterns, tool inventory, and decision trees that Claude's general training doesn't reliably cover — especially for fast-moving products like Couchbase.

Skills work with Claude Code, Claude.ai Projects, OpenAI Codex CLI, Cursor, Gemini CLI, and any agent that supports the open Agent Skills standard.


How to use these skills

Option 1 — Skills CLI (one command, works with Claude Code and most agents)

npx skills add celticht32/Couchbase-Skills-for-Claude.ai

This installs all skills into your project's .claude/skills/ directory. Claude Code picks them up automatically on the next session.

To install a single skill:

npx skills add celticht32/Couchbase-Skills-for-Claude.ai --skill couchbase-sqlpp-tuning

Option 2 — Claude Code (manual import via CLAUDE.md)

Clone the repo and add references to your project's CLAUDE.md:

@skills/couchbase/couchbase-mcp/SKILL.md
@skills/couchbase/couchbase-sqlpp-tuning/SKILL.md

Or load the full collection and let Claude route:

See @skills/ for available Couchbase skills. Load the relevant one before responding to any Couchbase question.

Option 3 — Claude.ai Projects

  1. Open claude.ai and create or open a Project
  2. Go to Project Knowledge → Add content
  3. Upload the SKILL.md files for the skills you want active
  4. Every conversation in that project will have them in context automatically

Option 4 — Any Claude interface (manual paste)

Open any skill's SKILL.md, copy the contents, and paste it into your system prompt or at the start of a conversation. For deep questions, also paste the relevant references/*.md file.


Best practices for Claude.ai Projects

A Claude.ai Project gives you a persistent Couchbase assistant — the right skills are always loaded, you never re-explain your environment, and every session starts with full context.

Recommended Project Knowledge

Don't load all 29 skills — that consumes most of your context window before you type anything. Load the skills you reach for in nearly every session and pull the others in as needed.

Always-on (upload these to Project Knowledge):

FileWhy
skills/couchbase/couchbase-mcp/SKILL.mdYou'll be calling MCP tools in almost every session
skills/couchbase/couchbase-sqlpp-tuning/SKILL.mdQuery questions come up constantly
skills/couchbase/couchbase-data-modeling/SKILL.mdDesign decisions arise early in any new work

Add reference files for your most common deep-dive topics. For example if you do a lot of query work, also upload:

  • skills/couchbase/couchbase-sqlpp-tuning/references/query-patterns.md
  • skills/couchbase/couchbase-sqlpp-tuning/references/explain-plan.md

For the other 26 skills: paste the relevant SKILL.md at the start of the conversation when you need it, or use @ imports in Claude Code.

Recommended Project Instructions

Set this in your Project's instruction field (the persistent system prompt for every conversation in the project):

I work with Couchbase clusters and MCP servers. Use the loaded skills before
answering any Couchbase question. When a question falls outside the loaded
skills, say so and I'll provide the relevant skill. Prefer concise answers.
Treat me as technical — skip basics.

Adjust to match your environment (self-managed vs Capella, your Couchbase version, your primary languages).

When to add more skills mid-conversation

If a session moves into territory not covered by your always-on skills — XDCR topology decisions, Kubernetes operator config, a migration plan — paste the relevant SKILL.md at that point:

"Here's the XDCR skill for this question: [paste contents of couchbase-xdcr/SKILL.md]"

Claude will use it for the rest of that conversation without it consuming Project Knowledge space permanently.


Skills

Core Couchbase (skills/couchbase/)

SkillCoverageRefs
couchbase-mcpOperating clusters via celticht32/MCP-Couchbase (167 tools) — KV, SQL++, FTS, indexes, XDCR, eventing, backup, KMIP, Capella v411
couchbase-app-integrationSDK integration — connection pooling, retry, durability, scan consistency, transactions, XDCR-aware patterns7
couchbase-data-modelingDocument design, boundaries, key formats, embed vs reference, TTL, time-series, vector modeling7
couchbase-sizingCluster sizing math — RAM, disk, node count, replicas, index memory budgets, Capella tier selection7
couchbase-sqlpp-tuningQuery tuning — EXPLAIN plans, index design, 15 named anti-patterns, CBO, array indexes, joins, pagination7
couchbase-migration-executionData migration — dual-write, CDC, cbimport, cutover runbooks, validation, rollback7
couchbase-ftsFull Text Search and vector search — index design, analyzers, query types, kNN, hybrid search, synonyms6
couchbase-ai-applicationsAI application design — HVI/CVI/SVI index selection, RAG pipelines, embedding strategies, LangChain/LlamaIndex4
couchbase-transactionsDistributed ACID transactions — when to use, two-phase commit, ATRs, retry, Python/Java/Node patterns3
couchbase-eventingEventing functions — OnUpdate/OnDelete, timers, curl(), N1QL in functions, deployment lifecycle3
couchbase-xdcrCross-datacenter replication — topology, active-active vs passive, conflict resolution, filtering4
couchbase-backup-restoreBackup and restore — cbbackupmgr, strategy, full/incremental, S3, point-in-time, Capella managed backups4
couchbase-mobileMobile / offline-first — Couchbase Lite, Sync Gateway, Capella App Services, sync functions3
couchbase-capellaCapella provisioning — getting started, networking (VPC peering, PrivateLink), database credentials3
couchbase-kubernetesKubernetes via Couchbase Autonomous Operator — CRDs, AZ awareness, rolling upgrades, PVs, Prometheus4
couchbase-security-hardeningProduction security — TLS, RBAC design, LDAP/SAML, audit logging, DARE encryption, network hardening6
couchbase-observabilityMonitoring — key metrics, Prometheus/Grafana, alert thresholds, log aggregation4
couchbase-performance-tuningCluster performance — KV latency, DCP backpressure, compaction, connection limits, OS tuning3
couchbase-upgradeVersion upgrades — upgrade paths, 8.0 breaking changes, Magma default, pre/post-upgrade checklist—
couchbase-magmaMagma storage engine — couchstore vs Magma, 128 vs 1024 vBuckets, compaction, memory requirements—
couchbase-columnarCapella Columnar — columnar analytics, links to operational data, BI tool integration2

Analytics Service (skills/couchbase-analytics/)

These skills document the celticht32/Couchbase-Analytics-MCP-Server.

SkillCoverage
cb-analytics-querySQL++ Analytics queries — tool selection, pagination, truncation, scan consistency
cb-analytics-schemaDataset discovery, schema inference, data dictionary workflows
cb-analytics-adminRuntime management — ingestion health, active requests, cancel, restart
cb-analytics-linksExternal data source links — S3, Azure Blob, GCS, remote Couchbase
cb-analytics-securityRBAC — users, groups, roles, service accounts, rate limits
cb-analytics-clusterCluster-level ops — nodes, memory quotas, rebalance, auto-failover, system events
cb-analytics-capellaCapella v4 control-plane — list/create/delete clusters, backups, restore
cb-analytics-mcp-setupInstalling and configuring the Analytics MCP server

Repository structure

skills/
  couchbase/           # 21 core Couchbase Server / Capella skills
    <skill-name>/
      SKILL.md
      references/
        <topic>.md
  couchbase-analytics/ # 8 Analytics MCP skills
    <skill-name>/
      SKILL.md

CLAUDE.md              # Claude Code project memory — contributor guidance
CONTRIBUTING.md        # Skill format spec and contribution workflow
README.md
LICENSE

Related projects

ProjectWhat it is
celticht32/MCP-Couchbase167-tool Couchbase MCP server documented by the couchbase-mcp skill
celticht32/Couchbase-Analytics-MCP-ServerAnalytics MCP server documented by the cb-analytics-* skills

Contributing

See CLAUDE.md for the skill format spec and CONTRIBUTING.md for the contribution workflow.


License

MIT — Copyright (c) 2026 Chris Ahrendt. See LICENSE.

Ver o original no GitHub

Bom para

Relacionados