convez.ai
SkillDevOps & InfraestruturaAvançado
meshmakers

OctoMesh Skills para Claude

meshmakers/octo-claude-skills

Automação de data mesh OctoMesh via linguagem natural no Claude

Ver no GitHub
InicialPluginScore · de 0 a 100

Conjunto de 10 skills Python que integram Claude ao OctoMesh, plataforma de data mesh. Permite orquestrar operações de CLI, pipelines ETL, deployments Kubernetes e desenvolvimento full-stack através de linguagem natural, com gates de confirmação para operações mutáveis.

Instalação

$npx convez-ai meshmakers/octo-claude-skills

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
10
Forks
0
Acompanhando
0
Saúde
37%
Último commit
há 2 dias
Versão
—
Issues abertas
1
Contribuidores
1

Casos de uso

  • Executar comandos octo-cli e explorar modelos GraphQL via conversa
  • Criar e validar pipelines ETL em YAML com semântica DataContext
  • Orquestrar builds, testes e deployments para test-2 sem UI
  • Desenvolver apps completos OctoMesh com CK model e HTTP APIs
  • Debugar falhas de blueprint, MongoDB e dependências NuGet

Veredito

Skill maduro e bem-estruturado para times que operam OctoMesh em escala. Requisitos rigorosos (octo-cli, PowerShell 7+, monorepo local) limitam seu alcance, mas execução é robusta com checkpoints de revisão antes de mutações. Recomendado para equipes já investidas no ecossistema.

README

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

octo-claude-skills

Claude Code plugin for the OctoMesh data mesh platform. Ten skills turn natural language into verified octo-cli commands, GraphQL exploration, build/devops operations, pipeline YAML, and development guidance — with confirmation gates for everything mutating.

Skills

SkillUse it for
/octoHub & CLI. Natural-language octo-cli (identity, tenants, blueprints, CK catalogs, stream-data archives, communication, AI services), CK model + runtime instance exploration via GraphQL, environment switching. Routes everything else to the right sibling skill.
/octo-devtoolsBuilding repos (Invoke-BuildAll), starting/stopping services (incl. the non-interactive agent pattern Start-Octo -nonInteractive $true / Stop-Octo), Docker + kind/Kubernetes infrastructure, git sync, NuGet, certificates.
/octo-agentDebugging and investigation: CK model failures (ResolveFailed, error code 66), MongoDB diagnostics, the DebugL NuGet dependency chain, selective builds, infrastructure backup/rollback, blueprint failures.
/octo-commitFinishing work: commit message format (`AB# New
/pipeline-expertETL pipeline YAML: authoring, node reference (SDK + Mesh Adapter + Zenon), DataContext semantics, validation against adapter schemas, debugging executions.
/octo-mcpDeveloping and extending the OctoMesh MCP server (octo-mcp-service): tool families, *ClientContext helpers, [McpRisk] classification, test conventions, adding-a-tool checklist.
/refinery-studioAngular development on the Data Refinery Studio (octo-frontend-refinery-studio): OctoGraphQlDataSource pattern, GraphQL codegen, LCARS theme system, lint/test gates.
/octo-operatorThe Kubernetes Communication Operator (octo-communication-operator): KubeOps, CommunicationPool CRDs, Helm values layering, TUnit/MTP testing, plus the octo-helm-core charts.
/octo-app-builderBuilding a complete OctoMesh-powered app: custom CK model → catalog, HTTP API as Mesh-Adapter pipelines, web UI as operator-deployed Application, packaged as one installable blueprint — the verified end-to-end recipe plus its pitfalls.
/octo-deployPromoting an app from local dev to the shared test-2 environment: shared-catalog publishing (octo-ckc/octo-bpm), docker.mm.cloud image push, test-2 context + tenant setup, install + kubectl-free verification. Staging/prod deliberately out of scope.

When in doubt, start with /octo — it routes to the right place.

Installation

# Add the marketplace and install the plugin
claude plugin marketplace add meshmakers/octo-claude-skills
claude plugin install octo-claude-skills

Prerequisites

  • octo-cli on PATH (local DebugL build or released version) with a configured context (~/.octo-cli/contexts.json)
  • PowerShell 7+ (pwsh) for the devtools skill — the OctoMesh workspace must contain octo-tools/
  • Python 3.x for the exploration scripts (venv is created automatically on first use)
  • The OctoMesh monorepo workspace for skills that read source (pipeline-expert, octo-agent, octo-mcp, refinery-studio, octo-operator)

Development

See CLAUDE.md for authoring conventions (description budgets, the 500-line SKILL.md rule, the accuracy mandate) and the versioning/release process. Validate after structural changes:

claude plugin validate . --strict

Release history lives in CHANGELOG.md.

Ver o original no GitHub

Bom para

Relacionados