golden-set-maintenance

golden-set-maintenance is a skill for Claude Code from latestaiagents/agent-skills. It costs 107 tokens per session (1,428 once invoked), scanned A, original, MIT.

A guide to maintaining a small set of high-value evaluation cases that must keep passing. This set acts as a fast check for important regressions, where a regression means a previously working behavior has broken.

In plain words
What is it for?
Use it to create a smoke check for every pull request, protect core user scenarios, or catch quality drops in safety-sensitive outputs.
Why use it?
It gives you a stable, quick signal when a full evaluation takes too long to run on every code change.

Skill for Claude Code

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

Part of the evals plugin — 5 skills shipped together , and of latestaiagents

Good fit Use it to create a smoke check for every pull request, protect core user scenarios, or catch quality drops in safety-sensitive outputs.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/latestaiagents/agent-skills/golden-set-maintenance
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 latestaiagents/agent-skills --skill golden-set-maintenance
Clone the repo
git clone --depth 1 https://github.com/latestaiagents/agent-skills

Made for: Claude Code.

Or install evals, the plugin that ships this one along with the rest of its 5 skills.

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 golden-set-maintenance

README.md
[![agentmods](https://agentmods.dev/badge/skills/latestaiagents/agent-skills/golden-set-maintenance/github.svg)](https://agentmods.dev/skills/latestaiagents/agent-skills/golden-set-maintenance)
Your own site
<a href="https://agentmods.dev/skills/latestaiagents/agent-skills/golden-set-maintenance"><img src="https://agentmods.dev/badge/skills/latestaiagents/agent-skills/golden-set-maintenance/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 golden-set-maintenance

Your own site · 80×15
<a href="https://agentmods.dev/skills/latestaiagents/agent-skills/golden-set-maintenance"><img src="https://agentmods.dev/badge/skills/latestaiagents/agent-skills/golden-set-maintenance.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 107 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,428 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 $0.00107 $0.01428
Opus 5 $0.00053 $0.00714
Sonnet 5 $0.00021 $0.00286
Haiku 4.5 $0.00011 $0.00143

Measured 7d ago against content hash d1b8e9ab77ea, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

golden-set-maintenance 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 7d 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/evals/golden-set-maintenance/SKILL.md · 153 lines

How it starts

The opening of the file, as written. The whole thing — 153 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Golden Set Maintenance

A golden set is 20-50 cases that matter most. If any fail, something important broke. Run them on every PR.

When to Use

  • Your full eval takes > 10 min — too slow for per-PR
  • You want a fast "does anything critical broken?" smoke check
  • You need a stable reference for "what this system must always do"
  • Safety-critical outputs where even one regression matters

What Goes In

High-signal cases ONLY. Each golden item should satisfy:

  1. Represents a core use case — if this fails, real users notice
  2. Unambiguous expected output — label is crisp, not subjective
  3. Has regressed at least once — historical anchor ("never again")
  4. Discriminating — different prompts/models yield different results

Reject:

  • "Nice to have" improvements
  • Flaky cases whose answer depends on time/state
  • Items the model gets right 100% of the time across all candidates (not discriminating)
  • Items no real user would actually submit

Size

  • Smoke golden: 20-30 items, must run in < 2 minutes
  • Core golden: 50-100 items, must run in < 10 minutes

Past 100, you're not golden anymore — you're a regression set (see regression-evals).

Curation Workflow

  1. Propose: anyone can add an item via PR. Include rationale ("this regressed in Oct 2025")
  2. Review: 2 reviewers verify the expected output is crisp and correct
  3. Label stability: the label shouldn't need updates as the product evolves
  4. Pass check: at least one model/prompt configuration should fail this case (otherwise not discriminating)

Selection Criteria Rubric

Before adding a golden item, answer:

- [ ] Is this a workflow real users actually do? (If no, don't add)
- [ ] Is the expected output objectively checkable? (If no, don't add)
- [ ] Would a 10% regression on this item be a P0 bug? (If no, don't add)
- [ ] Is there a similar item already in the set? (If yes, don't duplicate)
- [ ] Has a variation of this case failed before? (Bonus — strongly include)

Read the full file on GitHub · 153 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.

  1. 7d ago First seen · 153 lines · 107 tokens per session scan A d1b8e9ab77ea

Subscribe to this mod's changes

golden-set-maintenance is a skill published in the GitHub repository latestaiagents/agent-skills (5 stars, last pushed 4mo ago), licensed MIT. It adds 107 tokens to every session and 1,428 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