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 makifbaysal/tasktrooper --skill ml-testing-strategygit clone --depth 1 https://github.com/makifbaysal/tasktrooperWrote 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/makifbaysal/tasktrooper/ml-testing-strategy)<a href="https://agentmods.dev/skills/makifbaysal/tasktrooper/ml-testing-strategy"><img src="https://agentmods.dev/badge/skills/makifbaysal/tasktrooper/ml-testing-strategy/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/makifbaysal/tasktrooper/ml-testing-strategy"><img src="https://agentmods.dev/badge/skills/makifbaysal/tasktrooper/ml-testing-strategy.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 | $0.00053 | $0.01922 |
| Opus 5.5 | $0.00021 | $0.00769 |
| Sonnet 5.5 | $0.00011 | $0.00384 |
| Haiku 4.5 | $0.00005 | $0.00192 |
Grade A, and why
ml-testing-strategy 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 4d 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 — 154 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ML Testing Strategy
Overview
Data code is easy to "test" badly: a test that re-runs the pipeline and asserts the metric equals last week's 0.8312 breaks on every library upgrade and catches nothing. Good data tests pin behaviour on tiny, explicit fixtures where you know the right answer by hand, and assert model quality as a threshold on planted signal.
Core principle: test your logic, not the library; derive expected values by hand on a fixture small enough to read; assert ranges for anything learned.
1. The layers
| Layer | What it proves | Share of tests |
|---|---|---|
| Pure transforms | A cleaning rule, a feature, a metric gives the hand-derived answer | most |
| Schemas / contracts | Bad input is rejected with the right column named | one per rule |
| Properties (Hypothesis) | Invariants hold for many generated inputs | a few, on core transforms |
| Pipeline smoke | Entry point runs end-to-end on a fixture, writes a loadable artifact | one per entrypoint |
| Model quality | Beats the baseline by a margin on planted signal, fixed seed | one or two per model |
| Determinism and parity | Same seed → same output; saved/loaded and served predictions equal training-time ones | one each |
| Warehouse | dbt data and unit tests (sql-analytics-and-dbt) | per model |
2. Fixtures: small, synthetic, explicit
Build them in the test or conftest.py. Never a production extract (PII, size, drift) and never the network.
def test_order_value_sums_lines_and_ignores_cancelled() -> None:
lines = pd.DataFrame({
"order_id": ["o1", "o1", "o2", "o3"],
"amount": [10.0, 5.0, 7.0, 3.0],
"status": ["ok", "ok", "cancelled", "ok"],
})
expected = pd.Series({"o1": 15.0, "o3": 3.0}, name="order_value")
expected.index.name = "order_id"
pd.testing.assert_series_equal(order_value(lines), expected)
Every fixture for a transform carries the cases that break data code: a null, a duplicate key, an unseen category, an empty frame, a single row, a boundary timestamp, a timezone-aware value. Each earns its own named test.
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.
- 4d ago First seen · 154 lines · 53 tokens per session scan A 887e760cb8c6
ml-testing-strategy is a skill published in the GitHub repository makifbaysal/tasktrooper (113 stars, last pushed yesterday), licensed Apache-2.0. It adds 53 tokens to every session and 1,922 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 2026-10-09.
Other skills, from other repositories
nemo-automodel-model-onboarding
Guide for onboarding new model architectures into NeMo AutoModel, including architecture discovery, implementation patterns, registration, and validation.
nemo-automodel-recipe-development
Create and modify NeMo AutoModel training and evaluation recipes, including YAML structure, builders, and execution flow.
nemo-mbridge-mlm-bridge-training
Run Megatron-LM (MLM) and Megatron Bridge training with mock or real data. Covers correlation testing, available recipes, and multi-GPU examples.
agent-tester
Agent skill for tester - invoke with $agent-tester.
data-quality-frameworks
Implement data quality validation with Great Expectations, dbt tests, and data contracts. Use when building data quality pipelines, implementing validation rules, or establishing data contracts.
dbt-transformation-patterns
Master dbt (data build tool) for analytics engineering with model organization, testing, documentation, and incremental strategies. Use when building data transformations, creating data models, or implementing analytics engineering best practices.