eval-harness

eval-harness is a skill for Claude Code from mhylle/claude-skills-collection. It costs 135 tokens per session (2,502 once invoked), scanned A, original, MIT.

A toolkit for evaluating AI-assisted software, including capability checks, regression tests, automated graders, and measures such as accuracy, speed, cost, and pass rate. Evaluation-driven development means defining how success will be measured before or during implementation.

In plain words
What is it for?
Use it to build or run benchmarks, regression suites, code or model graders, and tracked evaluations for AI features, agents, or skills.
Why use it?
It turns vague claims about an AI feature into repeatable checks and makes regressions visible as the system changes.

Skill for Claude Code

Written for Claude Code: argument-hint in frontmatter.

Part of the devflow plugin — 38 skills, 13 agents, 5 hooks shipped together

Good fit Use it to build or run benchmarks, regression suites, code or model graders, and tracked evaluations for AI features, agents, or skills.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/mhylle/claude-skills-collection/eval-harness
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add mhylle/claude-skills-collection --skill eval-harness
Clone the repo
git clone --depth 1 https://github.com/mhylle/claude-skills-collection

Made for: Claude Code.

Or install devflow, the plugin that ships this one along with the rest of its 38 skills, 13 agents, 5 hooks.

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.

agentmods badge for eval-harness

README.md
[![agentmods](https://agentmods.dev/badge/skills/mhylle/claude-skills-collection/eval-harness/github.svg)](https://agentmods.dev/skills/mhylle/claude-skills-collection/eval-harness)
Your own site
<a href="https://agentmods.dev/skills/mhylle/claude-skills-collection/eval-harness"><img src="https://agentmods.dev/badge/skills/mhylle/claude-skills-collection/eval-harness/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.

agentmods 80×15 button for eval-harness

Your own site · 80×15
<a href="https://agentmods.dev/skills/mhylle/claude-skills-collection/eval-harness"><img src="https://agentmods.dev/badge/skills/mhylle/claude-skills-collection/eval-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 135 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,502 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. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00135 $0.02502
Opus 5 $0.00068 $0.01251
Sonnet 5 $0.00027 $0.00500
Haiku 4.5 $0.00014 $0.00250

Measured 11d ago against content hash 07f9dc21d42a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

eval-harness 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 11d 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.

skills/eval-harness/SKILL.md · 324 lines

How it starts

The opening of the file, as written. The whole thing — 324 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Eval Harness

A skill for building and running evaluation harnesses on AI features. Covers two eval types (capability, regression), three grader types (code, model, human), and the metrics that summarize them (pass@k, pass^k).


Evaluation-driven development (EDD)

Evals are defined before or alongside implementation, not after. Success criteria become explicit, measurable, and testable from day one.

Core principles:

  1. Define success first — before implementing a feature, define what "working correctly" means through explicit evaluations.
  2. Measure continuously — run evals throughout development, not just at the end.
  3. Automate where possible — prefer automated graders for speed and consistency.
  4. Human review for nuance — use human graders when quality judgments require context or subjectivity.
  5. Track regressions — every capability added becomes a regression test.
  6. Iterate on failures — failed evals provide specific, actionable feedback for improvement.

Benefits: clear success criteria (no ambiguity about what "done" means), early problem detection, confidence in changes, tests double as executable specifications, quantitative progress tracking.

The mindset shift:

Traditional: "Build it, then figure out if it works"
EDD:         "Define what 'works' means, then build to pass those evals"

Eval types (summary)

Type Purpose When to use
Capability Verify new functionality works Adding features, improving functionality, testing edge cases
Regression Protect against unintended breakage After code changes, before PR merge, after dependency updates

Full structures and YAML examples → references/eval-types.md.


Grader types (summary)

Type Strength When to use
Code (exact match, regex, function output) Fast, deterministic, cheap Any codifiable correctness check
Model (LLM judge with rubric) Handles semantic variation, tone, quality Natural-language outputs, multiple valid answers
Human (Likert, binary, ranking) Catches what machines miss Creative content, safety-critical judgments, calibrating auto-graders

Read the full file on GitHub · 324 lines

Files

What ships with it

4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

  1. 11d ago First seen · 324 lines · 135 tokens per session scan A 07f9dc21d42a

Subscribe to this mod's changes

eval-harness is a skill published in the GitHub repository mhylle/claude-skills-collection (18 stars, last pushed 8d ago), licensed MIT. It adds 135 tokens to every session and 2,502 once invoked, about $0.0007 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 2026-08-30.

Related

Other skills, from other repositories

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

jetson-validate-image

Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.

NVIDIA/skills · 50 tokens

atmos-validation

Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.

cloudposse/atmos · 31 tokens

skill-benchmark

Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.

HoangNguyen0403/agent-skills-standard · 16 tokens