Skip to main content
skill CX Codex

linkerd-expert — Claude Code skill: what it does, how to install

linkerd-expert — a Linkerd skill for Kubernetes work: the agent installs the mesh, injects the proxy, pulls linkerd viz metrics and writes access policies.

Apache-2.0 162 installs ★ 45 updated 2026-10-12

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add personamanagmentlayer/pcl --skill linkerd-expert -a claude-code
Open on GitHub

What linkerd-expert does

The skill hands the agent the Linkerd operator role on a Kubernetes cluster. It carries the CLI sequences: linkerd check --pre as a pre-flight, install and upgrade pipes into kubectl apply, and linkerd inject for putting the proxy onto a workload. Debugging runs through linkerd viz stat, tap, routes, edges and linkerd diagnostics proxy-metrics.

Injection is meant to be switched on per namespace with the linkerd.io/inject: enabled annotation, and proxy CPU and memory ceilings go in the same place. Retries arrive as an HTTPRoute filter, ServiceProfiles can be generated from an OpenAPI file, and ServerAuthorization rules deny by default. Common mistakes are called out too, like opening the whole inbound port range.

Linkerd 2.14 and Kubernetes 1.28 are the floor set for this skill, so anything older sits outside what it was written against. Shell work is narrowed to kubectl and linkerd; files are handled with Read, Write, Edit, Glob and Grep. Background material in references/CORE_CONCEPTS.md gets opened only when the task calls for it.

What it helps with

  • install Linkerd into a cluster and verify it with linkerd check
  • turn on proxy injection for one namespace
  • watch live requests between two deployments
  • build a ServiceProfile from an OpenAPI spec
  • lock a service down with ServerAuthorization

Example prompts for the agent

  • Run linkerd check --pre against my cluster and tell me what blocks the install.
  • Turn on injection for the production namespace and show me per-deployment success rates.
  • Tap traffic from deployment/myapp to deployment/api and explain where the latency comes from.

How to install linkerd-expert

1Skills CLI (Vercel)

npx skills add personamanagmentlayer/pcl --skill linkerd-expert -a claude-code

2Manually: put the files in place

~/.claude/skills/linkerd-expert/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/linkerd-expert/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about linkerd-expert

How to install linkerd-expert in Claude Code?

Run npx skills add personamanagmentlayer/pcl --skill linkerd-expert -a claude-code. The skill then lives at ~/.claude/skills/linkerd-expert/SKILL.md.

What does linkerd-expert do?

It guides the agent through Linkerd on Kubernetes: mesh install, proxy injection, traffic and metrics via linkerd viz, retries and access policies.

Which Linkerd and Kubernetes versions does it target?

The skill is written for Linkerd 2.14 and Kubernetes 1.28 or newer. It runs its commands through kubectl and linkerd.

Source text

SKILL.md · Apache-2.0
---
name: linkerd-expert
version: 1.1.0
description: >-
  Expert-level Linkerd service mesh management, traffic control, reliability, and
  production operations. Use when the user mentions service mesh, Kubernetes,
  microservices, mTLS, or observability, or when the task involves Linkerd Architecture,
  Mesh Injection, Traffic Management, or Reliability Features.
category: devops
author: PCL Team
license: Apache-2.0
tags:
  - linkerd
  - service-mesh
  - kubernetes
  - microservices
  - mtls
  - observability
allowed-tools:
  - Read
  - Write
  - Edit
  - Bash(kubectl:*, linkerd:*)
  - Glob
  - Grep
requirements:
  linkerd: '>=2.14'
  kubernetes: '>=1.28'
---

# Linkerd Expert

You are an expert in Linkerd service mesh with deep knowledge of traffic management, reliability features, security, observability, and production operations. You design and manage lightweight, secure microservices architectures using Linkerd's ultra-fast data plane.

## linkerd CLI Commands

**Installation and Status:**

```bash
# Pre-installation check
linkerd check --pre

# Install
linkerd install | kubectl apply -f -

# Check installation
linkerd check

# Upgrade
linkerd upgrade | kubectl apply -f -

# Uninstall
linkerd uninstall | kubectl delete -f -
```

**Mesh Operations:**

```bash
# Inject deployment
kubectl get deployment myapp -o yaml | linkerd inject - | kubectl apply -f -

# Inject namespace
linkerd inject deployment.yaml | kubectl apply -f -

# Uninject
linkerd uninject deployment.yaml | kubectl apply -f -
```

**Observability:**

```bash
# Stats
linkerd viz stat deployments -n production
linkerd viz stat pods -n production

# Routes
linkerd viz routes deployment/myapp -n production

# Top
linkerd viz top deployment/myapp -n production

# Tap (live traffic)
linkerd viz tap deployment/myapp -n production
linkerd viz tap deployment/myapp -n production --to deployment/api

# Edges (traffic graph)
linkerd viz edges deployment -n production
```

**Diagnostics:**

```bash
# Get proxy logs
linkerd viz logs deployment/myapp -n production

# Proxy metrics
linkerd viz metrics deployment/myapp -n production

# Diagnostics
linkerd diagnostics proxy-metrics pod/myapp-xxx -n production
```

## Best Practices

### 1. Use Automatic Injection

```yaml
# Enable at namespace level
annotations:
  linkerd.io/inject: enabled
```

### 2. Set Resource Limits

```yaml
annotations:
  config.linkerd.io/proxy-cpu-limit: '1000m'
  config.linkerd.io/proxy-memory-limit: '256Mi'
```

### 3. Configure Retries and Timeouts

```yaml
# Use HTTPRoute for reliability
filters:
  - type: RequestHeaderModifier
    requestHeaderModifier:
      set:
        - name: l5d-retry-limit
          value: '3'
```

### 4. Monitor Golden Metrics

```
- Success Rate (requests/sec)
- Request Volume (RPS)
- Latency (P50, P95, P99)
```

### 5. Use ServiceProfiles

```bash
# Generate from OpenAPI
linkerd viz profile myapp -n production --open-api swagger.json
```

### 6. Implement Zero Trust

```yaml
# Default deny, explicit allow
kind: ServerAuthorization
```

### 7. Multi-Cluster for HA

```bash
# Export critical services
mirror.linkerd.io/exported: "true"
```

## Anti-Patterns

**1. No Resource Limits:**

```yaml
# BAD: No proxy limits
# GOOD: Set explicit limits
config.linkerd.io/proxy-cpu-limit: '1000m'
```

**2. Skip Ports Unnecessarily:**

```yaml
# BAD: Skip all ports
config.linkerd.io/skip-inbound-ports: "1-65535"

# GOOD: Only skip specific ports (metrics, health)
config.linkerd.io/skip-inbound-ports: "9090"
```

**3. No Authorization Policies:**

```yaml
# GOOD: Always implement Server + ServerAuthorization
```

**4. Ignoring Metrics:**

```bash
# GOOD: Monitor success rate, latency, RPS
linkerd viz stat deployments -n production
```

## Approach

When implementing Linkerd:

1. **Start Simple**: Inject one service first
2. **Enable Namespace Injection**: Scale gradually
3. **Monitor**: Use viz dashboard and CLI
4. **Reliability**: Add retries and timeouts
5. **Security**: Implement authorization policies
6. **Profile Services**: Generate ServiceProfiles
7. **Multi-Cluster**: For high availability
8. **Tune**: Adjust proxy resources based on load

Always design service mesh configurations that are lightweight, secure, and observable following cloud-native principles.

## Reference Documentation

Detailed material lives alongside this skill and is read on demand:

- [Core Expertise](references/CORE_CONCEPTS.md) — Linkerd Architecture, Installation, Mesh Injection, Traffic Management, Reliability Features, Authorization Policies, Multi-Cluster, Observability

## Resources

- Linkerd Documentation: https://linkerd.io/docs/
- Linkerd Best Practices: https://linkerd.io/2/tasks/
- BuoyantCloud: https://buoyant.io/cloud
- Service Mesh Interface (SMI): https://smi-spec.io/

© PCL Team · Apache-2.0 · personamanagmentlayer/pcl

Similar skills

All skills for DevOps and infrastructure →

From the same set104

The whole set →
codeql-expert The codeql-expert skill guides CodeQL database setup, custom query writing, and vulnerability scans. Use it for SAST tasks, taint tracking, and GitHub security workflows. CX Codex +1 174installs skill-creator-expert skill-creator-expert guides you through designing, writing and checking PCL skills, from domain analysis to a frontmatter template and a quality checklist. CX Codex +1 173installs django-expert django-expert guides full-stack Django development covering ORM models, views, forms, authentication, and Django REST Framework APIs. CX Codex +1 172installs graphql-expert graphql-expert is a Claude Code skill for GraphQL schema design, resolvers, and subscriptions, with guidance on Apollo federation and N+1 queries. CX Codex +1 169installs data-mesh-expert data-mesh-expert walks the agent through data mesh work: domains own their data, products ship with contracts, governance is federated and the platform serves itself. CX Codex +1 168installs swift-expert swift-expert guides an agent through Swift work for iOS and macOS, from SwiftUI screens to async code. Use it when a task involves Combine or modern Swift features. CX Codex +1 163installs pcl-expert pcl-expert — a skill for Persona Control Language: persona, team and workflow declarations, the compiler pipeline, the pcl toolchain. Useful when the agent edits .pcl files or the compiler itself. CX Codex +1 160installs sql-expert sql-expert — a skill for expert SQL work: schema design, efficient queries, index optimization, and administration across PostgreSQL, MySQL, and SQL Server. CX Codex +1 1installs

Project sponsors

Companies that keep this analytics open