Agent Skills for AI coding assistants — teaches your agent the Conduktor platform.
Skills follow the Agent Skills format and work with Claude Code, Cursor, VS Code (Copilot), Gemini CLI, and 30+ other tools.
npx skills add conduktor/skillsConduktor platform expertise for Apache Kafka management, governance, and self-service. Covers Console, Gateway, and CLI.
Use when:
- Deploying Console and Gateway (Docker, Helm, Kubernetes)
- Configuring Gateway interceptors (encryption, data quality, traffic control)
- Setting up multi-tenancy with virtual clusters
- Automating with CLI, GitOps, or Terraform
- Onboarding application developers to Conduktor-managed Kafka
- Creating topics, requesting access, or connecting through Gateway
Covers two personas:
| Platform engineer | Application developer |
|---|---|
| Deploy Console & Gateway | Connect to Kafka through Gateway |
| Encrypt or mask Kafka data | Create topics via self-service |
| Enforce data quality rules | Produce & consume through Gateway |
| Set up multi-tenancy | Request access to other teams' topics |
| Apply traffic control & safeguards | |
| Automate with CLI & GitOps | |
| Manage infrastructure with Terraform |
Once installed, just ask. The agent doesn't just explain — it discovers your environment, asks the right questions, generates configs with real values, and executes.
Examples:
| You say | What the agent does |
|---|---|
| "Install Conduktor" | Checks Docker/Helm/kubectl, asks quick test vs existing cluster, deploys via Docker Compose or Helm on Kubernetes, polls health, sets up CLI auth |
| "Encrypt PII fields in my Kafka topics" | Discovers your interceptors and topics via CLI, asks which fields, generates EncryptPlugin YAML, offers conduktor apply |
| "Set up team isolation" | Lists existing virtual clusters, asks how many teams, generates VirtualCluster + ServiceAccount + Group YAML |
| "I need access to the payments topic" | Discovers topics and your ApplicationInstance, identifies the owner, generates the permission YAML |
| "Create a topic for my service" | Checks your AppInstance's TopicPolicy constraints, asks topic name and config, validates, applies |
| "Write Terraform for our Conduktor setup" | Exports current resources via CLI, generates matching HCL, offers terraform plan |
For details beyond what the skill files cover (specific interceptor configs, API endpoints, edge cases), Conduktor provides an MCP documentation server that agents can query in real time. Skills give static knowledge, the MCP server gives live lookups.
{
"mcpServers": {
"conduktor-docs": {
"type": "url",
"url": "https://docs.conduktor.io/mcp"
}
}
}This exposes a search_conduktor_documentation tool that searches across all Conduktor docs and returns contextual content with direct links.
skills/
└── conduktor/
├── SKILL.md # Entry point (auto-discovered)
├── references/
│ ├── mental-model.md # Product boundaries, resource model
│ ├── anti-patterns.md # Common AI assistant mistakes
│ └── terminology.md # Conduktor-to-Kafka mapping
└── use-cases/
├── platform/ # Platform engineer use cases
│ ├── deploy-conduktor.md
│ ├── encrypt-kafka-data.md
│ ├── enforce-data-quality.md
│ ├── multi-tenancy.md
│ ├── traffic-control.md
│ ├── gitops-automation.md
│ └── terraform.md
└── app-developer/ # Application developer use cases
├── onboard-to-kafka.md
├── create-topic.md
├── produce-consume.md
└── request-access.md