CtrlK
BlogDocsLog inGet started
Tessl Logo

ark-sdk-development

Regenerate and debug types across the ARK stack (SDK, API, Dashboard). Use when fixing TypeScript type errors in ark-dashboard, updating types after CRD changes, regenerating types.ts from OpenAPI spec, debugging "Property does not exist on type" schema errors, or adding custom SDK functionality via overlays. Covers the full type pipeline from Kubernetes CRDs to TypeScript.

97

1.05x
Quality

96%

Does it follow best practices?

Impact

100%

1.05x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

Evaluation results

100%

8%

Extending the ARK API with Workload Scheduling Models

Pydantic model domain-prefix naming

Criteria
Without context
With context

Consistent prefix used

100%

100%

No collision with agents.py names

100%

100%

No collision with mcp_servers.py names

100%

100%

No collision with queries.py names

100%

100%

Resource type defined

100%

100%

Priority class type defined

100%

100%

Affinity/selector type defined

100%

100%

Top-level schedule type defined

100%

100%

Naming rationale in design notes

100%

100%

Collision risk explained

42%

100%

100%

Adding a Metrics Client to the ARK SDK

SDK overlay custom utility placement

Criteria
Without context
With context

Correct overlay path

100%

100%

Not in generated directory

100%

100%

MetricsClient class present

100%

100%

Base URL on init

100%

100%

Counter increment method

100%

100%

Gauge record method

100%

100%

HTTP library used

100%

100%

Module docstring present

100%

100%

Correct build command

100%

100%

Overlay copy step explained

100%

100%

100%

7%

Fixing Dashboard Type Errors After API Expansion

TypeScript type error debugging and fix

Criteria
Without context
With context

Agents models renamed

100%

100%

MCP models renamed

83%

100%

Forward refs updated in agents

100%

100%

Forward refs updated in MCP

100%

100%

No __models__ names in TS

100%

100%

TS uses clean schema names

100%

100%

Root cause identified

100%

100%

Fix principle stated

100%

100%

No workaround used

100%

100%

Non-determinism or CI risk mentioned

50%

100%

Repository
mckinsey/agents-at-scale-ark
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.