eval-authoring

eval-authoring is a skill for Claude Code from sigistry/marketplace. It costs 96 tokens per session (957 once invoked), scanned A, original, MIT.

A method for testing changes to an AI model, prompt, or application with fixed examples and measurable checks. An evaluation, or eval, shows whether a change improved results or caused regressions.

In plain words
What is it for?
Use it to create test cases, choose fact-based or model-based graders, maintain golden datasets, and run regression checks in tools such as Promptfoo or DeepEval.
Why use it?
It replaces informal spot-checking with repeatable evidence and can stop a build when important behavior gets worse.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the llm-app-hardener plugin — 4 skills, 4 commands, 2 agents shipped together

Good fit Use it to create test cases, choose fact-based or model-based graders, maintain golden datasets, and run regression checks in tools such as Promptfoo or DeepEval.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/sigistry/marketplace/eval-authoring
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 sigistry/marketplace --skill eval-authoring
Clone the repo
git clone --depth 1 https://github.com/sigistry/marketplace

Made for: Claude Code.

Or install llm-app-hardener, the plugin that ships this one along with the rest of its 4 skills, 4 commands, 2 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.

agentmods badge for eval-authoring

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

Your own site · 80×15
<a href="https://agentmods.dev/skills/sigistry/marketplace/eval-authoring"><img src="https://agentmods.dev/badge/skills/sigistry/marketplace/eval-authoring.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 96 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 957 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00096 $0.00957
Opus 5 $0.00048 $0.00478
Sonnet 5 $0.00019 $0.00191
Haiku 4.5 $0.00010 $0.00096

Measured 6d ago against content hash 9c2cff8ed358, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

eval-authoring 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 6d 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/llm-app-hardener/skills/eval-authoring/SKILL.md · 40 lines

How it starts

The opening of the file, as written. The whole thing — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Eval Authoring

Purpose

Give LLM apps a repeatable way to escape "vibes-based" development, where a prompt is changed, a few outputs are eyeballed, and the change ships with no evidence it helped. An eval suite turns that into a measured, versioned, CI-gated check: every prompt or model change is scored against fixed cases, and a regression fails the build. This skill standardizes what to test, how to grade it, and how to wire it in, riding an existing framework (promptfoo, DeepEval, Ragas) when one is present, and falling back to plain unit tests plus an LLM judge when none is.

The three grader types

Grader Use for Cost / stability Example
Assertion Deterministic, checkable facts about the output Cheap, stable, run every commit is-json, json-schema, contains, regex, not-contains (banned strings), latency/cost budget
LLM-as-judge Qualities assertions can't capture Costs a call, needs a threshold Faithfulness, helpfulness, tone, "answers the question", pairwise vs. baseline
Golden dataset Regression over a curated set of real cases Grows from production traffic input → expected/assert, tagged, versioned in the repo

Prefer assertions wherever the property is checkable, they are free and non-flaky. Reach for a judge only for genuinely subjective qualities, and always give it a pass threshold. Use the golden set to lock in behavior you've already fixed so it never regresses.

Authoring workflow

  1. Name the contract. For the prompt/agent under test: its inputs, the output shape it promises, and the failure modes worth catching (wrong format, missing field, hallucination, ignored instruction, unsafe content, verbosity).
  2. Collect real cases. Pull 3–6 representative inputs from fixtures, logs, or the prompt's own examples. Never invent domain facts, a wrong "expected" answer is worse than no test.
  3. Layer the graders. Start with assertions for everything checkable; add judge tests for the rest; seed a golden file with the cases.
  4. Set thresholds. Every judge assertion needs a numeric pass bar. For regression gating, prefer pairwise (new output must be ≥ baseline) over an absolute score.
  5. Gate CI. Run the suite on PRs and fail on any assertion failure or judge score below threshold/baseline. Keep model-calling tests out of the every-commit path where possible.
  6. Grow the set. Every production failure becomes a new golden case, so the same bug can never ship twice.

Read the full file on GitHub · 40 lines

Files

What ships with it

2 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. 6d ago First seen · 40 lines · 96 tokens per session scan A 9c2cff8ed358

Subscribe to this mod's changes

eval-authoring is a skill published in the GitHub repository sigistry/marketplace (3 stars, last pushed 5d ago), licensed MIT. It adds 96 tokens to every session and 957 once invoked, about $0.0005 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 2026-09-03.

Related

Other skills, from other repositories

deep-plan

Creates detailed, sectionized, TDD-oriented implementation plans through research, stakeholder interviews, and multi-LLM review. Use when planning features that need thorough pre-implementation analysis.

piercelamb/deep-plan · 39 tokens

tdd

Test-driven development — the red-green-refactor discipline for code that's correct by design, not by accident. INVOKE PROACTIVELY whenever writing or changing code, fixing a bug, adding or structuring tests, or choosing what to mock — even when nobody says "TDD" or "tests". (Tests are code held to…

PrabhdeepSingh/claude-plugins · 80 tokens

wave-execution-framework-v2

Use when executing multi-wave engineering work needing strict TDD, bug-capture/fix split, quality gates, and orchestrated teams with per-agent Opus-advisor / Sonnet-executor model routing.

Harshvardhan86/claude-wave-plugin · 48 tokens

nano-banana

This skill should be used for Python scripting and Gemini image generation. Use when users ask to generate images, create AI art, edit images with AI, or run Python scripts with uv. Trigger phrases include "generate an image", "create a picture", "draw", "make an image of", "nano banana", or any image generation…

NikiforovAll/claude-code-rules · 73 tokens

dotnet-verify

This skill should be used when working with Verify snapshot tests in .NET projects. Use when updating verified snapshots after intentional code changes, accepting new snapshots, discovering verify tests, or troubleshooting snapshot mismatches. Trigger phrases include "verify tests", "update snapshots", "accept…

NikiforovAll/claude-code-rules · 68 tokens

auto-loop

TDD-based autonomous development loop with checkpoint recovery and observability changelog.

claude-world/director-mode-lite · 17 tokens