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 cosmix/loom --skill loom-testinggit clone --depth 1 https://github.com/cosmix/loomWrote 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/cosmix/loom/loom-testing)<a href="https://agentmods.dev/skills/cosmix/loom/loom-testing"><img src="https://agentmods.dev/badge/skills/cosmix/loom/loom-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/cosmix/loom/loom-testing"><img src="https://agentmods.dev/badge/skills/cosmix/loom/loom-testing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 96 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00025 | $0.02747 |
| Opus 5.5 | $0.00010 | $0.01099 |
| Sonnet 5.5 | $0.00005 | $0.00549 |
| Haiku 4.5 | $0.00003 | $0.00275 |
Grade A, and why
loom-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 1mo 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 — 236 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Testing
Overview
Writing tests: unit/integration/e2e plus data-pipeline, ML, and infrastructure domains. This file owns test-double taxonomy, AAA, and framework patterns. For pyramid ratios, coverage targets, risk prioritization, and flaky-test diagnosis, see loom-test-strategy; for browser/Playwright/Cypress, see loom-e2e-testing.
Workflow
- Map the unit — public interface, dependencies/side effects, invariants, error paths, boundary values.
- Pick the altitude — unit for logic/branches; integration for real collaborators (DB, HTTP) at a boundary; e2e for user journeys. Push detail down the pyramid (
loom-test-strategy). - Write failing test first when practical (TDD red→green→refactor): a red test proves the test actually exercises the code; a test that never failed asserts nothing.
- Arrange-Act-Assert, one logical assertion per test, deterministic inputs.
AAA and naming
- Arrange state/doubles → Act (one call to the unit) → Assert outcome. Blank-line separate the three; more than one Act means split the test.
- Name
test_<unit>_<scenario>_<expected>— e.g.test_cart_add_duplicate_increases_quantity. The name is the spec; a failing name should tell you what broke without reading the body. - One logical assert per test. Multiple physical asserts on one behavior are fine (
status, thenbody); asserting two unrelated behaviors is "assertion roulette" — the first failure masks the rest. Prefer one composite assert (assert_eq!(got, expected_struct)) over many field asserts.
Test Doubles — taxonomy and misuse
Precision matters: "mock" is colloquially any double, but the kind you choose decides whether the test is brittle.
| Double | Does | Verifies | Reach for when |
|---|---|---|---|
| Dummy | Fills a parameter, never used | nothing | Satisfy a signature |
| Stub | Returns canned values for indirect inputs | state (result) | Drive a branch from a dependency's return |
| Spy | Stub that records how it was called | state + calls, after | Assert an effect happened, loosely |
| Mock | Pre-set call expectations, self-verifying | interaction, strict | The interaction is the contract (e.g. "charge called once with amount") |
| Fake | Lightweight working impl (in-memory DB, fake clock) | state | Collaborator too slow/awkward for real, but behavior matters |
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.
- 1mo ago First seen · 236 lines · 25 tokens per session scan A 84efd8785021
loom-testing is a skill published in the GitHub repository cosmix/loom (62 stars, last pushed today), licensed MIT. It adds 25 tokens to every session and 2,747 once invoked, about $0.0001 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 2026-09-03.
Other skills, from other repositories
langchain4j-testing-strategies
Provides unit test, integration test, and mock AI patterns for LangChain4j applications. Creates mock LLM responses, tests retrieval chains, validates RAG workflows, and implements Testcontainers-based integration tests for Java AI services. Use when unit testing AI services, integration testing LangChain4j…
continuous-testing
Continuous test-driven development loop — after every code change, builds the project, starts the server, and runs Unit Tests, Integration Tests, and System Tests. Applies on top of microprofile-server skill. Use during development when you want full verification after each change. Triggers on "continuous testing"…
Test Strategy
How to decide what to test at what level — the test pyramid, risk-based coverage, and matching test depth to change risk.
Android Testing
JUnit + Robolectric unit tests, Compose UI tests, screenshot tests, and Espresso for legacy views.
iOS Testing
XCTest unit tests, XCUITest UI tests, snapshot testing, and performance baselines for iOS apps.
Spring Test Pyramid
Unit, integration, slice, and E2E test patterns for Spring Boot applications.