Patterns and techniques for evaluating and improving AI agent outputs.
日本語の概要は準備中です。原文の説明を表示しています。
Deploys .NET containers. Kubernetes probes, Docker Compose for local dev, CI/CD integration.
インストール方法を見るインストールする前に、エージェントに与えられる指示の中身を確認できます。
Deploying .NET containers to Kubernetes and local development environments. Covers Kubernetes Deployment + Service + probe YAML, Docker Compose for local dev workflows, and CI/CD integration for building and pushing container images.
Cross-references: [skill:dotnet-containers] for Dockerfile and image best practices, [skill:dotnet-observability] for health check endpoint patterns used by Kubernetes probes.
A production-ready Kubernetes Deployment for a .NET API:
apiVersion: apps/v1
kind: Deployment
metadata:
name: order-api
labels:
app: order-api
app.kubernetes.io/name: order-api
app.kubernetes.io/version: '1.0.0'
app.kubernetes.io/component: api
spec:
replicas: 3
selector:
matchLabels:
app: order-api
template:
metadata:
labels:
app: order-api
spec:
containers:
- name: order-api
image: ghcr.io/myorg/order-api:1.0.0
ports:
- containerPort: 8080
protocol: TCP
env:
- name: ASPNETCORE_ENVIRONMENT
value: 'Production'
- name: OTEL_EXPORTER_OTLP_ENDPOINT
value: 'http://otel-collector.monitoring:4317'
- name: OTEL_SERVICE_NAME
value: 'order-api'
- name: ConnectionStrings__DefaultConnection
valueFrom:
secretKeyRef:
name: order-api-secrets
key: connection-string
resources:
requests:
cpu: '100m'
memory: '128Mi'
limits:
cpu: '500m'
memory: '512Mi'
livenessProbe:
httpGet:
path: /health/live
port: 8080
initialDelaySeconds: 10
periodSeconds: 15
timeoutSeconds: 3
failureThreshold: 3
readinessProbe:
httpGet:
path: /health/ready
port: 8080
initialDelaySeconds: 5
periodSeconds: 10
timeoutSeconds: 3
failureThreshold: 3
startupProbe:
httpGet:
path: /health/live
port: 8080
initialDelaySeconds: 0
periodSeconds: 5
failureThreshold: 30
securityContext:
runAsNonRoot: true
runAsUser: 1654
fsGroup: 1654
terminationGracePeriodSeconds: 30
```text
### Service Manifest
Expose the Deployment within the cluster:
```yaml
apiVersion: v1
kind: Service
metadata:
name: order-api
labels:
app: order-api
spec:
type: ClusterIP
selector:
app: order-api
ports:
- port: 80
targetPort: 8080
protocol: TCP
name: http
```text
### ConfigMap for Non-Sensitive Configuration
```yaml
apiVersion: v1
kind: ConfigMap
metadata:
name: order-api-config
data:
ASPNETCORE_ENVIRONMENT: 'Production'
Logging__LogLevel__Default: 'Information'
Logging__LogLevel__Microsoft.AspNetCore: 'Warning'
```text
Reference in the Deployment:
```yaml
envFrom:
- configMapRef:
name: order-api-config
```yaml
### Secrets for Sensitive Configuration
```yaml
apiVersion: v1
kind: Secret
metadata:
name: order-api-secrets
type: Opaque
stringData:
connection-string: 'Host=postgres;Database=orders;Username=app;Password=<DB_PASSWORD_PLACEHOLDER>'
```text
In production, use an external secrets operator (e.g., External Secrets Operator, Sealed Secrets) rather than plain
Kubernetes Secrets stored in source control.
---
## Kubernetes Probes
Probes tell Kubernetes how to check application health. They map to the health check endpoints defined in your .NET
application (see [skill:dotnet-observability]).
### Probe Types
| Probe | Purpose | Endpoint | Failure Action |
| ------------- | ---------------------------------- | --------------- | ------------------------------------------------------- |
| **Startup** | Has the app finished initializing? | `/health/live` | Keep waiting (up to `failureThreshold * periodSeconds`) |
| **Liveness** | Is the process healthy? | `/health/live` | Restart the pod |
| **Readiness** | Can the process serve traffic? | `/health/ready` | Remove from Service endpoints |
### Probe Configuration Guidelines
```yaml
# Startup probe: give the app time to initialize
# Total startup budget: failureThreshold * periodSeconds = 30 * 5 = 150s
startupProbe:
httpGet:
path: /health/live
port: 8080
initialDelaySeconds: 0
periodSeconds: 5
failureThreshold: 30
# Liveness probe: detect deadlocks and hangs
# Only runs after startup probe succeeds
livenessProbe:
httpGet:
path: /health/live
port: 8080
periodSeconds: 15
timeoutSeconds: 3
failureThreshold: 3
# Readiness probe: control traffic routing
readinessProbe:
httpGet:
path: /health/ready
port: 8080
periodSeconds: 10
timeoutSeconds: 3
failureThreshold: 3
```text
### Graceful Shutdown
.NET responds to `SIGTERM` and begins graceful shutdown. Configure `terminationGracePeriodSeconds` to allow in-flight
requests to complete:
```yaml
spec:
terminationGracePeriodSeconds: 30
```yaml
In your application, use `IHostApplicationLifetime` to handle shutdown:
```csharp
app.Lifetime.ApplicationStopping.Register(() =>
{
// Perform cleanup: flush telemetry, close connections
Log.CloseAndFlush();
});
```text
Ensure the `Host.ShutdownTimeout` allows in-flight requests to complete:
```csharp
builder.Host.ConfigureHostOptions(options =>
{
options.ShutdownTimeout = TimeSpan.FromSeconds(25);
});
```text
Set `ShutdownTimeout` to a value less than `terminationGracePeriodSeconds` to ensure the app shuts down before
Kubernetes sends `SIGKILL`.
---
## Docker Compose for Local Development
Docker Compose provides a local development environment that mirrors production dependencies.
### Basic Compose File
```yaml
# docker-compose.yml
services:
order-api:
build:
context: .
dockerfile: src/OrderApi/Dockerfile
ports:
- '8080:8080'
environment:
- ASPNETCORE_ENVIRONMENT=Development
- ConnectionStrings__DefaultConnection=Host=postgres;Database=orders;Username=app;Password=<DB_PASSWORD_PLACEHOLDER>
- OTEL_EXPORTER_OTLP_ENDPOINT=http://otel-collector:4317
depends_on:
postgres:
condition: service_healthy
redis:
condition: service_healthy
# Note: CMD-SHELL + curl requires a base image with shell and curl installed.
# Chiseled/distroless images lack both. For chiseled images, either use a
# non-chiseled dev target in the Dockerfile or omit the healthcheck and rely
# on depends_on ordering (acceptable for local dev).
healthcheck:
test: ['CMD-SHELL', 'curl -f http://localhost:8080/health/live || exit 1']
interval: 10s
timeout: 3s
retries: 3
start_period: 10s
postgres:
image: postgres:17
environment:
POSTGRES_DB: orders
POSTGRES_USER: app
POSTGRES_PASSWORD: devpassword
ports:
- '5432:5432'
volumes:
- postgres-data:/var/lib/postgresql/data
healthcheck:
test: ['CMD-SHELL', 'pg_isready -U app -d orders']
interval: 5s
timeout: 3s
retries: 5
redis:
image: redis:7-alpine
ports:
- '6379:6379'
healthcheck:
test: ['CMD', 'redis-cli', 'ping']
interval: 5s
timeout: 3s
retries: 5
volumes:
postgres-data:
```text
### Development Override
Use a separate override file for development-specific settings:
```yaml
# docker-compose.override.yml (auto-loaded by docker compose up)
services:
order-api:
build:
target: build # Stop at build stage for faster rebuilds
volumes:
- .:/src # Mount source for hot reload
environment:
- ASPNETCORE_ENVIRONMENT=Development
- DOTNET_USE_POLLING_FILE_WATCHER=true
command: ['dotnet', 'watch', 'run', '--project', 'src/OrderApi/OrderApi.csproj']
```bash
### Observability Stack
Add an OpenTelemetry collector and Grafana for local observability:
```yaml
# docker-compose.observability.yml
services:
otel-collector:
image: otel/opentelemetry-collector-contrib:latest
command: ['--config=/etc/otelcol-config.yaml']
volumes:
- ./infra/otelcol-config.yaml:/etc/otelcol-config.yaml
ports:
- '4317:4317' # OTLP gRPC
- '4318:4318' # OTLP HTTP
grafana:
image: grafana/grafana:latest
ports:
- '3000:3000'
volumes:
- grafana-data:/var/lib/grafana
volumes:
grafana-data:
```text
Run with the observability stack:
```bash
docker compose -f docker-compose.yml -f docker-compose.observability.yml up
```bash
---
## CI/CD Integration
Basic CI/CD patterns for building and pushing .NET container images. Advanced CI patterns (matrix builds, environment
promotion, deploy pipelines) -- see [skill:dotnet-gha-publish], [skill:dotnet-gha-deploy], and
[skill:dotnet-ado-publish].
### GitHub Actions: Build and Push
```yaml
# .github/workflows/docker-publish.yml
name: Build and Push Container
on:
push:
branches: [main]
tags: ['v*']
env:
REGISTRY: ghcr.io
IMAGE_NAME: ${{ github.repository }}
jobs:
build-and-push:
runs-on: ubuntu-latest
permissions:
contents: read
packages: write
steps:
- uses: actions/checkout@v4
- name: Log in to container registry
uses: docker/login-action@v3
with:
registry: ${{ env.REGISTRY }}
username: ${{ github.actor }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Extract metadata
id: meta
uses: docker/metadata-action@v5
with:
images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
tags: |
type=ref,event=branch
type=semver,pattern={{version}}
type=semver,pattern={{major}}.{{minor}}
type=sha
- name: Build and push
uses: docker/build-push-action@v6
with:
context: .
push: true
tags: ${{ steps.meta.outputs.tags }}
labels: ${{ steps.meta.outputs.labels }}
cache-from: type=gha
cache-to: type=gha,mode=max
```text
### Image Tagging Strategy
| Tag Pattern | Example | Use Case |
| ----------- | ------------------- | ------------------------------------------- |
| `latest` | `myapi:latest` | Development only -- never use in production |
| Semver | `myapi:1.2.3` | Release versions -- immutable |
| Major.Minor | `myapi:1.2` | Floating tag for patch updates |
| SHA | `myapi:sha-abc1234` | Unique per commit -- traceability |
| Branch | `myapi:main` | CI builds -- latest from branch |
### dotnet publish Container in CI
For projects using `dotnet publish /t:PublishContainer` instead of Dockerfiles:
```yaml
steps:
- uses: actions/checkout@v4
- uses: actions/setup-dotnet@v4
with:
dotnet-version: '10.0.x'
- name: Publish container image
run: |
dotnet publish src/OrderApi/OrderApi.csproj \
--os linux --arch x64 \
/t:PublishContainer \
-p:ContainerRegistry=${{ env.REGISTRY }} \
-p:ContainerRepository=${{ env.IMAGE_NAME }} \
-p:ContainerImageTag=${{ github.sha }}
```text
---
## Key Principles
- **Use startup probes** to decouple initialization time from liveness detection -- without a startup probe,
slow-starting apps get killed before they are ready
- **Separate liveness from readiness** -- liveness checks should not include dependency health (see
[skill:dotnet-observability] for endpoint patterns)
- **Set resource requests and limits** -- without them, pods can starve other workloads or get OOM-killed unpredictably
- **Run as non-root** -- set `runAsNonRoot: true` in the pod security context and use chiseled images (see
[skill:dotnet-containers])
- **Use `depends_on` with health checks** in Docker Compose -- prevents app startup before dependencies are ready
- **Keep secrets out of manifests** -- use Kubernetes Secrets with external secrets operators, not plain values in
source control
- **Match ShutdownTimeout to terminationGracePeriodSeconds** -- ensure the app finishes cleanup before Kubernetes sends
SIGKILL
---
## Agent Gotchas
1. **Do not omit the startup probe** -- without it, the liveness probe runs during initialization and may restart
slow-starting apps. Calculate startup budget as `failureThreshold * periodSeconds`.
2. **Do not include dependency checks in liveness probes** -- a database outage should not restart your app. Liveness
endpoints must only check the process itself. See [skill:dotnet-observability] for the liveness vs readiness pattern.
3. **Do not use `latest` tag in Kubernetes manifests** -- `latest` is mutable and `imagePullPolicy: IfNotPresent` may
serve stale images. Use immutable tags (semver or SHA).
4. **Do not hardcode connection strings in Kubernetes manifests** -- use Secrets or ConfigMaps referenced via
`secretKeyRef`/`configMapRef`.
5. **Do not set `terminationGracePeriodSeconds` lower than `Host.ShutdownTimeout`** -- the app needs time to drain
in-flight requests before Kubernetes sends SIGKILL.
6. **Do not forget `condition: service_healthy` in Docker Compose `depends_on`** -- without the condition, Compose
starts dependent services immediately without waiting for health checks.
---
## Code Navigation (Serena MCP)
**Primary approach:** Use Serena symbol operations for efficient code navigation:
1. **Find definitions**: `serena_find_symbol` instead of text search
2. **Understand structure**: `serena_get_symbols_overview` for file organization
3. **Track references**: `serena_find_referencing_symbols` for impact analysis
4. **Precise edits**: `serena_replace_symbol_body` for clean modifications
**When to use Serena vs traditional tools:**
- ✅ **Use Serena**: Navigation, refactoring, dependency analysis, precise edits
- ✅ **Use Read/Grep**: Reading full files, pattern matching, simple text operations
- ✅ **Fallback**: If Serena unavailable, traditional tools work fine
**Example workflow:**
```text
# Instead of:
Read: src/Services/OrderService.cs
Grep: "public void ProcessOrder"
# Use:
serena_find_symbol: "OrderService/ProcessOrder"
serena_get_symbols_overview: "src/Services/OrderService.cs"
```
## References
- [Deploy ASP.NET Core to Kubernetes](https://learn.microsoft.com/en-us/aspnet/core/host-and-deploy/linux-nginx)
- [Kubernetes Deployments](https://kubernetes.io/docs/concepts/workloads/controllers/deployment/)
- [Kubernetes probes](https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/)
- [Docker Compose overview](https://docs.docker.com/compose/)
- [ASP.NET Core health checks](https://learn.microsoft.com/en-us/aspnet/core/host-and-deploy/health-checks)
- [Graceful shutdown in .NET](https://learn.microsoft.com/en-us/dotnet/core/extensions/generic-host#host-shutdown)
- [GitHub Actions: Publishing Docker images](https://docs.github.com/en/actions/use-cases-and-examples/publishing-packages/publishing-docker-images)
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Patterns and techniques for evaluating and improving AI agent outputs.
日本語の概要は準備中です。原文の説明を表示しています。
Comprehensive AI prompt engineering safety review and improvement prompt. Analyzes prompts for safety, bias, security vulnerabilities, and effectiveness while providing detailed improvement recommendations.
日本語の概要は準備中です。原文の説明を表示しています。
Use when user requests research requiring multiple sources, comprehensive analysis, or synthesis across topics - technical research, domain knowledge gathering, market analysis, or learning about complex subjects
日本語の概要は準備中です。原文の説明を表示しています。
AI-powered wiki generation for code repositories with commands, agents, and skills
日本語の概要は準備中です。原文の説明を表示しています。
Use when building .NET 10 or C# 14 applications; when using minimal APIs, modular monolith patterns, or feature folders; when implementing HTTP resilience, Options pattern, Channels, or validation; when seeing outdated patterns like old extension method syntax
日本語の概要は準備中です。原文の説明を表示しています。
Implements accessible .NET UI. SemanticProperties, ARIA, AutomationPeer, testing per platform.
日本語の概要は準備中です。原文の説明を表示しています。