A generator for language-model evaluation tests for MCP tool collections. These tests check workflows through the Claude Agent SDK, rather than testing each handler directly.

Use it after building tool collections and compiling the project to create evaluation setup and scenario files under tests/evals/. It can generate tests for all collections or one named collection.

It separates probabilistic AI behavior checks from deterministic integration tests. This helps reveal whether related tools work together for realistic tasks.

build-evals — Skill for Claude Code by umbraco

Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo

git clone --depth 1 https://github.com/umbraco/Umbraco-MCP-Base

agentmods

npx agentmods add skills/umbraco/umbraco-mcp-base/build-evals

Made for: Claude Code.

Or install umbraco-mcp-skills, the plugin that ships this one along with the rest of its 13 skills, 7 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/umbraco/umbraco-mcp-base/build-evals/github.svg)](https://agentmods.dev/skills/umbraco/umbraco-mcp-base/build-evals)

Your own site

<a href="https://agentmods.dev/skills/umbraco/umbraco-mcp-base/build-evals"><img src="https://agentmods.dev/badge/skills/umbraco/umbraco-mcp-base/build-evals/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/umbraco/umbraco-mcp-base/build-evals"><img src="https://agentmods.dev/badge/skills/umbraco/umbraco-mcp-base/build-evals.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 36 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 3,279 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00036	$0.03279
Opus 5	$0.00018	$0.01639
Sonnet 5	$0.00007	$0.00656
Haiku 4.5	$0.00004	$0.00328

Measured 12d ago against content hash e31b6e0a5fc1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

build-evals scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/umbraco-mcp-skills/skills/build-evals/SKILL.md · 379 lines

How it starts

The opening of the file, as written. The whole thing — 379 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Build Evals

Generate LLM eval tests for MCP tool collections created by /build-tools. This skill reads .discover.json and the existing tool files, then builds eval test files in tests/evals/ one collection at a time.

IMPORTANT: This skill ONLY creates files inside tests/evals/ — the setup file and eval test files. Do NOT create or modify tool files, collection indexes, integration tests, mock handlers, or any other files.

Key difference from integration tests (/build-tools-tests):

Integration tests: one file per tool, test handlers directly against real API, deterministic
Eval tests: group related tools by workflow, test via Claude Agent SDK, probabilistic, require npm run build first (run against dist/)

Prerequisites

Before running, ensure:

You have run /build-tools (tool collections exist in src/umbraco-api/tools/)
The project builds: npm run build
An LLM API key is available (eval tests use Claude Agent SDK — works with a Claude Code subscription or ANTHROPIC_API_KEY)

Arguments

No arguments: build evals for all collections from .discover.json
Single collection name: build evals only for that collection (e.g. /build-evals form)

Agents

This skill orchestrates the following agent — use it for the relevant step:

Agent	When to use
`eval-test-creator`	Creating eval test files (Step 5)

Critical Rules

BUILD BEFORE RUNNING. Eval tests run against dist/index.js. Always npm run build first.

ONE COLLECTION AT A TIME. Complete each collection before starting the next.

GROUP TOOLS BY WORKFLOW. Unlike integration tests (one file per tool), eval tests group related tools into workflow scenarios. A collection with 5 tools might have just 1-2 eval test files.

ITERATE ON PROMPTS. Eval tests are probabilistic. If a test fails, the fix is usually in the prompt — make instructions more explicit, add search steps for IDs, use unique identifiers.

VERBOSE DURING DEVELOPMENT. Always set verbose: true when creating/debugging. Disable after tests pass reliably.

Read the full file on GitHub · 379 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

12d ago First seen · 379 lines · 36 tokens per session scan A e31b6e0a5fc1

Subscribe to this mod's changes

build-evals is a skill published in the GitHub repository umbraco/Umbraco-MCP-Base (2 stars, last pushed yesterday), licensed MIT. It adds 36 tokens to every session and 3,279 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Testing

research-engineer

An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.

davila7/claude-code-templates · 43 tokens

tika-eval-compare

Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".

apache/tika · 50 tokens

neuron-evaluation-engineer

Create and run AI evaluations with datasets, assertions, and output drivers in Neuron AI. Use this skill whenever the user mentions evaluation, testing AI systems, creating evaluators, dataset-driven testing, assertion-based validation, or wants to measure AI system performance. Also trigger for tasks involving…

neuron-core/neuron-ai · 77 tokens

build-evals