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 agentmods add skills/hoavdc/codexkit/codexkit-a-b-test-plannernpx skills add hoavdc/CodexKit --skill codexkit-a-b-test-plannergit clone --depth 1 https://github.com/hoavdc/CodexKitWhat 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 | $0.00062 | $0.01366 |
| Opus 5 | $0.00031 | $0.00683 |
| Sonnet 5 | $0.00012 | $0.00273 |
| Haiku 4.5 | $0.00006 | $0.00137 |
Grade A, and why
codexkit-a-b-test-planner 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 2d 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 — 167 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A/B Test Planner
When to Use
- Before running any product experiment or feature test
- When optimizing conversion funnels or UX flows
- When leadership requires statistical rigor for feature decisions
- When planning multi-variant tests or sequential experiments
Procedure
Step 1 — Hypothesis
Write a clear, falsifiable hypothesis:
- If [change we're making]
- Then [metric we expect to change]
- Because [reasoning / user insight]
Step 2 — Metrics
| Type | Metric | Current Baseline |
|---|---|---|
| Primary | The metric that determines success | [value] |
| Secondary | Supporting metrics that provide context | [value] |
| Guardrail | Metrics that must NOT degrade | [value] |
Step 3 — Sample Size & Duration
Calculate required sample size using:
- Baseline conversion rate (p₁)
- Minimum Detectable Effect (MDE) — smallest meaningful change
- Statistical significance level (α) — typically 0.05
- Statistical power (1−β) — typically 0.80
n = f(p₁, MDE, α, β) → use standard sample size calculator
Duration = n / (daily traffic × allocation %)
Step 4 — Randomization Plan
- Randomization unit: user, session, device, or account
- Allocation: 50/50, or asymmetric with justification
- Stratification: any segments to balance (geography, plan, device)
- Exclusion: users to exclude (employees, bots, existing tests)
Step 5 — Decision Rules
| Outcome | Criteria | Action |
|---|---|---|
| Winner | Primary metric ↑ ≥ MDE, p < 0.05, guardrails stable | Ship to 100% |
| Neutral | No significant difference | Keep control, iterate hypothesis |
| Loser | Primary metric ↓ significantly | Revert, analyze why |
| Guardrail breach | Any guardrail metric degrades > threshold | Stop test immediately |
Step 6 — Rollout Playbook
- Ramp: 5% → 25% → 50% → 100% over [days]
- Monitoring: check metrics daily during ramp
- Rollback trigger: guardrail breach or unexpected anomaly
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.
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.
- 2d ago First seen · 167 lines · 62 tokens per session scan A d7d9039d7dde
codexkit-a-b-test-planner is a skill published in the GitHub repository hoavdc/CodexKit (21 stars, last pushed 3mo ago), licensed MIT. It adds 62 tokens to every session and 1,366 once invoked, about $0.0003 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
reducing-aigc-detection
Systematically reduce AIGC detection rates in academic papers (Chinese/English). Analyzes detection reports, identifies high-impact sections, applies multi-layer rewriting strategies preserving formatting/footnotes, and verifies results. Supports 维普/知网/Turnitin platforms.
architecting-security
安全架构与治理：威胁建模 (STRIDE/PASTA/LINDDUN)、零信任身份架构、IAM/SSO/MFA/PAM、合规框架 (SOC2/PCI/HIPAA/GDPR)、DLP、隐私工程、安全控制设计。Use when designing security architecture, threat modeling new systems, implementing zero-trust identity, designing IAM/SSO/PAM, building compliance evidence chains, or planning privacy-by-design.
defending-applications
Application security defense knowledge for builders. Covers Web/API/GraphQL hardening (XSS/SQLi/SSRF/IDOR/BOLA/Mass Assignment/deserialization/upload/path traversal), authentication/authorization (OAuth 2.0/OIDC/JWT/Session/Cookie/SAML/SSO), and LLM application security (prompt injection, jailbreak, RAG poisoning…
detecting-and-responding
蓝队与紫队工程：检测规则编写、SIEM/EDR 调优、事件响应、数字取证、威胁狩猎、ATT&CK 映射、紫队演练闭环。Use when writing Sigma/YARA detection rules, tuning SIEM noise, responding to security incidents, conducting forensic analysis, hunting threats, or running purple team exercises.
securing-cloud-and-supply-chain
云原生与软件供应链安全防御。容器/K8s 加固、Service Mesh、CI/CD 安全、SLSA/SBOM/Sigstore、云 IAM、Secrets 管理、IaC 安全。Use when hardening Kubernetes clusters, auditing CI/CD pipelines, implementing supply chain security, managing cloud IAM, or reviewing IaC code.
verification-loop
Evidence-before-assertions workflow. Use before claiming work is done, before release, and after any behavior change in scripts/skills/MCP.