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.
npx skills add Infinite-Labs-AI/infinite-skills --skill ab-testinggit clone --depth 1 https://github.com/Infinite-Labs-AI/infinite-skillsWrote 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.
[](https://agentmods.dev/skills/infinite-labs-ai/infinite-skills/ab-testing)<a href="https://agentmods.dev/skills/infinite-labs-ai/infinite-skills/ab-testing"><img src="https://agentmods.dev/badge/skills/infinite-labs-ai/infinite-skills/ab-testing/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.
<a href="https://agentmods.dev/skills/infinite-labs-ai/infinite-skills/ab-testing"><img src="https://agentmods.dev/badge/skills/infinite-labs-ai/infinite-skills/ab-testing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00032 | $0.00579 |
| Opus 5 | $0.00016 | $0.00290 |
| Sonnet 5 | $0.00006 | $0.00116 |
| Haiku 4.5 | $0.00003 | $0.00058 |
Grade A, and why
ab-testing 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.
How it starts
The opening of the file, as written. The whole thing — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A/B Testing
Turn a growth idea into a test that can actually change a decision.
Frame The Decision
Start with the decision the experiment should inform:
- Ship, kill, iterate, scale, or investigate.
- Audience or surface being tested.
- Current baseline.
- Primary metric and guardrail metric.
- Minimum effect that would matter.
- Sample size or traffic reality.
- Time window and implementation cost.
If the traffic is too low for an A/B test, recommend a qualitative, sequential, or directional test instead.
Write The Hypothesis
Use this shape:
Because [observed problem], changing [specific thing] for [audience] should improve [primary metric] without hurting [guardrail], shown by [measurement].
Make the variant isolate one main idea. Do not mix headline, price, layout, offer, and audience changes unless the test is explicitly a bundled concept test.
Choose The Test Type
Pick the method based on traffic, risk, and decision cost:
- A/B test: enough traffic and a reversible surface.
- Before/after read: operational change where randomization is impractical.
- Concierge test: validate demand or workflow manually before building.
- Smoke test: test interest before full fulfillment.
- Fake-door test: measure intent when the feature or offer is not ready, with ethical disclosure.
- Qualitative read: use interviews, session reviews, or sales calls when numbers will be too thin.
Add decision economics:
- Cost of shipping the wrong thing.
- Cost of waiting.
- Minimum useful evidence.
Design The Test
Define:
- Control and variant.
- Inclusion and exclusion rules.
- Primary metric.
- Guardrails.
- Instrumentation requirements.
- Decision threshold.
- Stop conditions.
- Rollback plan.
Interpret Carefully
- Do not call a winner before the decision threshold is met.
- Do not ignore novelty effects.
- Segment after the primary read, not until a desired story appears.
- Treat inconclusive results as useful when they eliminate bad ideas.
What ships with it
1 file 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.
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.
- 11d ago First seen · 110 lines · 32 tokens per session scan A de1c98bcf8d8
ab-testing is a skill published in the GitHub repository Infinite-Labs-AI/infinite-skills (44 stars, last pushed 12d ago), licensed MIT. It adds 32 tokens to every session and 579 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 2026-08-30.
Other skills, from other repositories
marketing-growth
Use when planning product marketing, writing marketing copy, designing pricing tiers, building programmatic SEO pages, optimizing on-page SEO, planning social media calendars, running Twitter automation, conducting Google Trends research, brainstorming brand/domain names, or producing YouTube videos.
generate-social-media-captions
Draft platform-appropriate social captions using a specific audience, message, voice, and call to action.
ad-creative-agent
A guide for editing one user-supplied photo, such as changing its background, removing objects, retouching it, or applying a visual style.
media-narrative-analysis
An article and media-message analysis skill for examining news reports, public-relations copy, competitor communications, policy notices, and social-media posts.
follow-loop
A manually triggered research workflow that gathers evidence about selected technology and luxury topics, then sends verified findings to long-term records.
spoken-content-extractor
A skill for extracting and reusing content from podcasts, livestreams, talks, interviews, and their recordings or transcripts. It can turn spoken material into organised written resources.