CtrlK
BlogDocsLog inGet started
Tessl Logo

vertex-ai-endpoint-config

Vertex Ai Endpoint Config - Auto-activating skill for GCP Skills. Triggers on: vertex ai endpoint config, vertex ai endpoint config Part of the GCP Skills skill category.

36

1.03x

Quality

3%

Does it follow best practices?

Impact

99%

1.03x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./planned-skills/generated/14-gcp-skills/vertex-ai-endpoint-config/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

100%

Deploy a Custom ML Model to Vertex AI

gcloud-based model deployment

Criteria
Without context
With context

Uses gcloud for model upload

100%

100%

Uses gcloud for endpoint creation

100%

100%

Uses gcloud for model deployment

100%

100%

No Python SDK usage

100%

100%

Correct project and region flags

100%

100%

Machine type specified

100%

100%

Minimum replica count set

100%

100%

Artifact URI referenced

100%

100%

Step-by-step guide produced

100%

100%

Prerequisites documented

100%

100%

Without context: $0.2613 · 1m 27s · 11 turns · 12 in / 5,004 out tokens

With context: $0.4674 · 1m 59s · 25 turns · 24 in / 6,782 out tokens

100%

Configure A/B Traffic Splitting on a Vertex AI Endpoint

endpoint traffic management

Criteria
Without context
With context

Uses gcloud for traffic update

100%

100%

No Python SDK usage

100%

100%

Traffic split flag used

100%

100%

Traffic percentages sum to 100

100%

100%

Conservative v2 allocation

100%

100%

Correct endpoint and project flags

100%

100%

Verification step included

100%

100%

Rollout strategy explained

100%

100%

Step-by-step structure

100%

100%

Without context: $0.1711 · 52s · 12 turns · 13 in / 2,817 out tokens

With context: $0.3626 · 1m 30s · 23 turns · 178 in / 4,401 out tokens

97%

7%

Set Up a Scalable Vertex AI Endpoint for a High-Traffic NLP Service

production endpoint scaling configuration

Criteria
Without context
With context

Uses gcloud for endpoint creation

100%

100%

Uses gcloud for model deployment

100%

100%

No Python SDK usage

100%

100%

Machine type specified

100%

100%

Minimum replica count set

0%

70%

Maximum replica count set

100%

100%

Model resource name referenced

100%

100%

Endpoint display name set

100%

100%

Validation checklist uses gcloud

100%

100%

Checklist has multiple checks

100%

100%

Without context: $0.3080 · 1m 12s · 17 turns · 18 in / 4,227 out tokens

With context: $0.3446 · 1m 31s · 21 turns · 19 in / 5,524 out tokens

Repository
jeremylongshore/claude-code-plugins-plus-skills
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.