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 joao-aires/agent-skills-public --skill ai-testinggit clone --depth 1 https://github.com/joao-aires/agent-skills-publicWrote 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/joao-aires/agent-skills-public/ai-testing)<a href="https://agentmods.dev/skills/joao-aires/agent-skills-public/ai-testing"><img src="https://agentmods.dev/badge/skills/joao-aires/agent-skills-public/ai-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/joao-aires/agent-skills-public/ai-testing"><img src="https://agentmods.dev/badge/skills/joao-aires/agent-skills-public/ai-testing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00048 | $0.01124 |
| Opus 5.5 | $0.00019 | $0.00450 |
| Sonnet 5.5 | $0.00010 | $0.00225 |
| Haiku 4.5 | $0.00005 | $0.00112 |
Grade A, and why
ai-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 yesterday.
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 — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Test AI behavior at the relevant boundaries
Start from the accepted user task, architecture and consequential failure modes. This method owns what to evaluate and what the evidence means; it does not select an evaluation SDK, require a model for a protocol-only service or authorize live calls and data publication. Reuse sufficient existing criteria and tests rather than restarting evaluation design for every change.
Define acceptance and evidence levels
| Property | Acceptance method |
|---|---|
| Deterministic invariants | Assert authorization, resource isolation, schema validity, side effects and recovery in the application, even when a model judge scores the answer highly. |
| Uncertain model quality | Define task-specific criteria for grounding, ambiguity, refusal and relevant safety constraints; choose suitable measures and graders. |
State provisional thresholds and uncertainty when acceptance is unresolved; do not invent product commitments.
Read evaluation methods when choosing levels, cases or graders. Assess model/retrieval components, protocol/runtime integration, agent trajectories, conversations, multimodal interaction and complete user journeys where the architecture needs them. Use actual compatible peers and persistent service state for integration and outcome claims. A component score, simulated user or recorded output establishes a narrower boundary than a live end-to-end interaction.
Build representative cases and suitable measures
Select cases
Derive cases from the task the application promises to perform. Include inputs on which different implementations could make consequentially different choices, and specify what would make each result useful or unacceptable. A case should reveal something about the capability; a large collection of interchangeable easy prompts can give a high score while leaving the difficult behavior unexamined.
Use meaningful user/data slices and privacy-reviewed observed failures to identify variation. Keep cases used while developing the behavior separate from held-out cases used to assess whether the improvement generalizes.
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.
- yesterday Changed · +32 lines 8bc9a30d2ab4
- 2d ago First seen · 42 lines · 48 tokens per session scan A 5bb938ec6242
ai-testing is a skill published in the GitHub repository joao-aires/agent-skills-public (2 stars, last pushed yesterday), licensed MIT. It adds 48 tokens to every session and 1,124 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-10-20.
Other skills, from other repositories
cli-eval
Create and run evaluation suites, watch live benchmark progress, view scorecards, compare model performance, and integrate eval runs with CI workflows from the CLI.
model-merging
Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or experimenting rapidly with model variants. Covers SLERP, TIES-Merging, DARE, Task…
darwinian-evolver
Evolve prompts/regex/SQL/code with Imbue's evolution loop.
dummy-dataset
Generate realistic dummy datasets for testing with customizable columns, constraints, and output formats (CSV, JSON, SQL, Python script). Use when creating test data, building mock datasets, or generating sample data for development and demos.
validate
Validate Semantica pipelines, extraction quality, graph schemas, and ontology consistency. Returns structured error/warning checklists. Uses PipelineValidator, PipelineBuilder.validatepipeline(), GraphValidator, and OntologyValidator. Sub-commands: pipeline, step, dependencies, extraction, graph, ontology, performance.
bad-case-to-dpo
A guide to turning failed production agent runs into training and evaluation data. It explains how to find the first mistake, create preference examples for DPO (Direct Preference Optimization), and check that a fix does not damage other tasks.