kubernetes

Name: kubernetes
Author: mjunaidca

22installs11forks

How to install kubernetes?

mjunaidca/mjs-agent-skills/kubernetes

$npx -y skills add mjunaidca/mjs-agent-skills --skill kubernetes

Installs into the current project.

Summary

Production-grade Kubernetes manifests and debugging for containerized applications. This skill should be used when users ask to deploy to Kubernetes, create K8s manifests, containerize for K8s, set up Deployments/Services/Jobs/StatefulSets/CronJobs, create namespaces with resourc

Files · 1

View on GitHub

SKILL.md

1	# Kubernetes
2
3	Production-grade K8s manifests with security-first defaults and educational comments.
4
5	---
6
7	## Resource Detection & Adaptation
8
9	Before generating manifests, detect the target environment:
10
11	```bash
12	# Detect node resources
13	kubectl get nodes -o jsonpath='{range .items[*]}{.metadata.name}: {.status.capacity.memory}, {.status.capacity.cpu}{"\n"}{end}'
14
15	# Detect if Docker Desktop (local) or real cluster
16	kubectl get nodes -o jsonpath='{.items[0].metadata.labels.node\.kubernetes\.io/instance-type}' 2>/dev/null \|\| echo "local"
17
18	# Detect available resources
19	kubectl describe nodes \| grep -A 5 "Allocated resources"
20	```
21
22	Adapt configurations based on detection:
23
24	\| Detected Environment \| Profile \| Default Limits \| Agent Action \|
25	\|---------------------\|---------\|----------------\|--------------\|
26	\| Docker Desktop < 6GB \| Minimal \| 128Mi-256Mi \| Warn, reduce replicas \|
27	\| Docker Desktop 6-10GB \| Standard \| 256Mi-512Mi \| Normal deployment \|
28	\| Cloud/Real cluster \| Production \| Based on node size \| Full features \|
29
30	### Agent Behavior
31
32	1. Detect cluster type and resources before generating manifests
33	2. Adapt resource requests/limits to cluster capacity
34	3. Warn if requested workload exceeds available resources
35	4. Calculate safe limits: `(node_memory * 0.7) / expected_pod_count`
36
37	### Adaptive Resource Templates
38
39	Local/Constrained (< 6GB allocatable):
40	```yaml
41	resources:
42	requests:
43	memory: 128Mi
44	cpu: 100m
45	limits:
46	memory: 256Mi
47	cpu: 500m
48	```
49
50	Standard (6-16GB allocatable):
51	```yaml
52	resources:
53	requests:
54	memory: 256Mi
55	cpu: 100m
56	limits:
57	memory: 512Mi
58	cpu: 1000m
59	```
60
61	Production (> 16GB or cloud):
62	```yaml
63	resources:
64	requests:
65	memory: 512Mi
66	cpu: 250m
67	limits:
68	memory: 1Gi
69	cpu: 2000m
70	```
71
72	### Pre-Deployment Validation
73
74	Before applying manifests, agent should verify:
75	```bash
76	# Check if deployment would exceed node capacity
77	kubectl get nodes -o jsonpath='{.items[0].status.allocatable.memory}'
78	```
79
80	If insufficient: warn user and suggest scaling down or increasing Docker Desktop resources.
81
82	---
83
84	## What This Skill Does
85
86	Analysis & Detection:
87	- Auto-detects from Dockerfile: ports, health endpoints, resources
88	- Identifies workload type from project structure
89	- Reads existing manifests to understand patterns
90	- Detects GPU requirements from dependencies
91
92	Generation:
93	- Creates production-hardened manifests (non-root, read-only, resource limits)
94	- Generates all supporting resources (Service, ConfigMap, HPA, PDB)
95	- Creates namespace governance (ResourceQuota, LimitRange, NetworkPolicy)
96	- Supports multi-team isolation with environment progression (dev → staging → prod)
97	- Adds educational comments explaining WHY each config choice
98	- Outputs ArgoCD-compatible directory structure
99
100	Validation:
101	- Verifies kubectl context exists
102	- Creates namespace if needed
103	- Deploys to local cluster (kind/minikube)
104	- Confirms pods are running before delivering
105
106	Security:
107	- Non-root user by default (runAsNonRoot: true)
108	- Read-only root filesystem
109	- No privilege escalation
110	- Dropped capabilities
111	- Resource limits always set
112	- Unprivileged ports only (>=1024) - privileged ports (<1024) require root
113
114	## What This Skill Does NOT Do
115
116	- Generate Helm charts (document in references for future)
117	- Create Kustomize overlays (document in references for future)
118	- Handle Dapr sidecar injection (separate skill)
119	- Deploy Kafka/Strimzi operators (separate skill)
120	- Generate ArgoCD Application CRDs (separate skill)
121
122	---
123
124	## Before Implementation
125
126	Gather context to ensure successful implementation:
127
128	\| Source \| Gather \|
129	\|--------\|--------\|
130	\| Codebase \| Dockerfile, existing manifests, port/health patterns \|
131	\| Conversation \| Target environment, namespace, special requirements \|
132	\| Skill References \| Security contexts, health probes, resource limits \|
133	\| User Guidelines \| Cluster conventions, naming standards \|
134
135	---
136
137	## Required Clarifications
138
139	After auto-detection, confirm with user if ambiguous:
140
141	\| Question \| When to Ask \|
142	\|----------\|-------------\|
143	\| Target environme

People who liked this also viewed

6 picks

Type

kubernetes

22installs11forks

View on GitHub

How to install kubernetes?

mjunaidca/mjs-agent-skills/kubernetes

$npx -y skills add mjunaidca/mjs-agent-skills --skill kubernetes

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Kubernetes
2
3	Production-grade K8s manifests with security-first defaults and educational comments.
4
5	---
6
7	## Resource Detection & Adaptation
8
9	Before generating manifests, detect the target environment:
10
11	```bash
12	# Detect node resources
13	kubectl get nodes -o jsonpath='{range .items[*]}{.metadata.name}: {.status.capacity.memory}, {.status.capacity.cpu}{"\n"}{end}'
14
15	# Detect if Docker Desktop (local) or real cluster
16	kubectl get nodes -o jsonpath='{.items[0].metadata.labels.node\.kubernetes\.io/instance-type}' 2>/dev/null \|\| echo "local"
17
18	# Detect available resources
19	kubectl describe nodes \| grep -A 5 "Allocated resources"
20	```
21
22	Adapt configurations based on detection:
23
24	\| Detected Environment \| Profile \| Default Limits \| Agent Action \|
25	\|---------------------\|---------\|----------------\|--------------\|
26	\| Docker Desktop < 6GB \| Minimal \| 128Mi-256Mi \| Warn, reduce replicas \|
27	\| Docker Desktop 6-10GB \| Standard \| 256Mi-512Mi \| Normal deployment \|
28	\| Cloud/Real cluster \| Production \| Based on node size \| Full features \|
29
30	### Agent Behavior
31
32	1. Detect cluster type and resources before generating manifests
33	2. Adapt resource requests/limits to cluster capacity
34	3. Warn if requested workload exceeds available resources
35	4. Calculate safe limits: `(node_memory * 0.7) / expected_pod_count`
36
37	### Adaptive Resource Templates
38
39	Local/Constrained (< 6GB allocatable):
40	```yaml
41	resources:
42	requests:
43	memory: 128Mi
44	cpu: 100m
45	limits:
46	memory: 256Mi
47	cpu: 500m
48	```
49
50	Standard (6-16GB allocatable):
51	```yaml
52	resources:
53	requests:
54	memory: 256Mi
55	cpu: 100m
56	limits:
57	memory: 512Mi
58	cpu: 1000m
59	```
60
61	Production (> 16GB or cloud):
62	```yaml
63	resources:
64	requests:
65	memory: 512Mi
66	cpu: 250m
67	limits:
68	memory: 1Gi
69	cpu: 2000m
70	```
71
72	### Pre-Deployment Validation
73
74	Before applying manifests, agent should verify:
75	```bash
76	# Check if deployment would exceed node capacity
77	kubectl get nodes -o jsonpath='{.items[0].status.allocatable.memory}'
78	```
79
80	If insufficient: warn user and suggest scaling down or increasing Docker Desktop resources.
81
82	---
83
84	## What This Skill Does
85
86	Analysis & Detection:
87	- Auto-detects from Dockerfile: ports, health endpoints, resources
88	- Identifies workload type from project structure
89	- Reads existing manifests to understand patterns
90	- Detects GPU requirements from dependencies
91
92	Generation:
93	- Creates production-hardened manifests (non-root, read-only, resource limits)
94	- Generates all supporting resources (Service, ConfigMap, HPA, PDB)
95	- Creates namespace governance (ResourceQuota, LimitRange, NetworkPolicy)
96	- Supports multi-team isolation with environment progression (dev → staging → prod)
97	- Adds educational comments explaining WHY each config choice
98	- Outputs ArgoCD-compatible directory structure
99
100	Validation:
101	- Verifies kubectl context exists
102	- Creates namespace if needed
103	- Deploys to local cluster (kind/minikube)
104	- Confirms pods are running before delivering
105
106	Security:
107	- Non-root user by default (runAsNonRoot: true)
108	- Read-only root filesystem
109	- No privilege escalation
110	- Dropped capabilities
111	- Resource limits always set
112	- Unprivileged ports only (>=1024) - privileged ports (<1024) require root
113
114	## What This Skill Does NOT Do
115
116	- Generate Helm charts (document in references for future)
117	- Create Kustomize overlays (document in references for future)
118	- Handle Dapr sidecar injection (separate skill)
119	- Deploy Kafka/Strimzi operators (separate skill)
120	- Generate ArgoCD Application CRDs (separate skill)
121
122	---
123
124	## Before Implementation
125
126	Gather context to ensure successful implementation:
127
128	\| Source \| Gather \|
129	\|--------\|--------\|
130	\| Codebase \| Dockerfile, existing manifests, port/health patterns \|
131	\| Conversation \| Target environment, namespace, special requirements \|
132	\| Skill References \| Security contexts, health probes, resource limits \|
133	\| User Guidelines \| Cluster conventions, naming standards \|
134
135	---
136
137	## Required Clarifications
138
139	After auto-detection, confirm with user if ambiguous:
140
141	\| Question \| When to Ask \|
142	\|----------\|-------------\|
143	\| Target environme

People who liked this also viewed

6 picks

Type