testing
Skill0 starsChanged 7 days agoMITView on GitHub
Use when writing, reviewing or running backend tests in any language pack - unit, table or parametrised, integration, fuzz, benchmark, race-policy and regression tests, or deciding how to test a change without sleeps or real time.
194 other projects use a version of this.
Adopted by
Most recently taken up by
- FAHTHIZ/oasistheir own version, first seen Nov 16, 2026
- fatihkan/baditheir own version, first seen Nov 16, 2026
- magnusrodseth/dotfilestheir own version, first seen Nov 16, 2026
- skillberry-ai/cap-evolvetheir own version, first seen Nov 15, 2026
- NVIDIA/nvcftheir own version, first seen Nov 15, 2026
- adelrodriguez/skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use bundle-dependency-prs, update-library-template and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
MicrosoftSkillbundle-dependency-prs
Fix broken dependency update PRs and aggregate the ones that work into one PR.
Also in 7 projects that use this
MicrosoftSkillupdate-library-template
Merges the latest Library.Template into this repo (at position of HEAD) and resolves conflicts.
Also in 7 projects that use this
zetaloopSkillassign-copilot
Assigns a GitHub issue to the Copilot coding agent, optionally specifying a custom agent. Use this when asked to assign an issue to Copilot or delegate an issue to CCA.
Also in 3 projects that use this
NVIDIA-NeMoSkillbuild-and-dependency
Dev environment setup for Megatron Bridge — container-based development, uv package management, lockfile regeneration, adding dependencies, Slurm container usage, and common build pitfalls.
Also in 3 projects that use this
timmo001Skillchangeset-scope
Keep all scoped code work contained to the user-defined changeset. Use for implementation, fixes, diagnosis, refactoring, cleanup, and review when explicit instructions, named files, diffs, branches, pull requests, or injected work scopes define the boundary.
Also in 3 projects that use this
JetBrainsSkillmodule-dependencies
Add or modify IntelliJ module dependencies in `.iml` files.
Also in 3 projects that use this
Other versions194 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 194
- modelcontextprotocol/servers★ 92ksame name
- lobehub/lobe-chat★ 86ksame name
- lobehub/lobehub★ 86ksame name
- penpot/penpot★ 62ksame name
- meteor/meteor★ 45ksame name
- iOfficeAI/AionUi★ 34ksame name
- garrytan/gbrain★ 33ksame name
- OpenZeppelin/openzeppelin-contracts★ 27ksame name
- tursodatabase/turso★ 25ksame name
- redis/go-redis★ 22ksame name
184 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository35 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- skills/work/platforms/claude.mdCLAUDE.md
- llms.txtllms.txt
- clarifySkill
- consistencySkill
- debuggingSkill
- documentationSkill
- featureSkill
- hexagonalSkill
- initSkill
- jobs-and-state-machinesSkill
- lang-clojureSkill
- lang-cppSkill
- lang-csharpSkill
- lang-elixirSkill
- lang-erlangSkill
- lang-goSkill
- lang-javaSkill
- lang-kotlinSkill
- lang-phpSkill
- lang-pythonSkill
- lang-rubySkill
- lang-rustSkill
- lang-scalaSkill
- lang-swiftSkill
- lang-typescriptSkill
- memorySkill
- parallel-workSkill
- performanceSkill
- qualitySkill
- reportSkill
- review-integrationSkill
- statusSkill
- workSkill
The file
We split this skill into the rules it gives an agent
1 rule. 0 are common, 0 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 174 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 34
Use when many inputs share one behaviour and setup (Go tables, pytest parametrize, JUnit @ParameterizedTest).
“Common” is measured: kept by at least 6 in 10 of the 174 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Testing and benchmarking
Load
- Your language:
skills/lang-<id>/testing.mdandconcurrency.md(loadloops:lang-<id>),specs/languages/<id>/RULES.mdTEST section. Go:specs/languages/go/RULES.md §30testing,§31benchmarks. specs/ARCHITECTURE.md §24testing architecture.- Gate checks and the race policy:
specs/WORKFLOW.md §6.
Always the pack's race policy
Every test run follows the pack's race policy (WORKFLOW §6): the race flag (flag packs: Go -race, Swift --sanitize=thread), the race build (build: the C++ TSan preset), or the substitute checks (none: randomised order with a printed seed, parallel runs, repeats of the changed tests). loops lang show <id> shows race. The answer to a slow suite is cheaper tests, never a weaker policy: no retries, no fixed order. Agents check their scope with loops gate --only <check>,… --pkgs <owned>; the full suite is the lead's and QA's (ROLES.md).
Pyramid
- Unit tests for pure rules (domain).
- Application tests with fake outbound ports.
- Adapter tests per technology: the framework's in-process test client for handlers (Go
httptest), a real database (container) for repositories, a fake server for client contract tests. - Integration tests across real boundaries only where needed.
- End-to-end checks against the running service (the product review).
- Fuzz tests for parsers, decoders, migrations, validators and quantity/money arithmetic.
- Benchmarks for hot paths; soak tests for leaks and invariants where the service is long-running.
Behaviour, not trivia
A test that would still pass with every function under test returning nothing is hollow: QA fails it (specs/ENGINEERING.md §12), and the quality pass's mutation testing finds it (loops:quality). See every new test fail for the reason you expect before you make it pass (§7). Name tests for behaviour (refund rejects an already refunded order; Go TestRefund_RejectsAlreadyRefundedOrder), not tests tied to private helper call counts. One reason to fail per test.
Table and parametrised tests
Use when many inputs share one behaviour and setup (Go tables, pytest parametrize, JUnit @ParameterizedTest). Do not force rows with unrelated setup into a table. Each case is named; parallel cases share no mutable state.
No sleeps, injected time
Inject the Clock; advance it in the test. Wait on completion signals (channels, futures, latches) or Eventually helpers with a deadline, never a sleep. Randomness comes from a seeded source. Tests must not depend on map order, wall clock, ports in use or the network.
Headless
Domain and application tests need no network, database, filesystem or listener. Config and fixture validators run without starting the service.
Fuzz and properties
Fuzz round trips (encode/decode, migrate old to new), parsers and validators; assert invariants (no panic, bounded output, idempotent apply). Commit interesting corpus entries where the pack's fuzzer reads them (Go: testdata/fuzz).
Benchmarks
Representative scale (N realistic, not toy). Report allocations where the harness can; separate setup from timed work; keep results alive so the work is not optimised away. Compare runs statistically (skills/performance; the harness per pack is in skills/lang-<id>/testing.md).
No-flap and consistency tests
Threshold-driven behaviour ships a no-flap test (skills/jobs-and-state-machines). Consistency-sensitive changes add repeat-run, duplicate-delivery and migration-fixture tests (skills/consistency).
Regression rule
For a bug: reproduce; add a failing test or fixture; fix; verify the test passes; keep the regression artifact. Never weaken or delete a test to get green; if a test is wrong, say so in the claims report.
Coverage (hard gate)
New code ships with unit tests. loops gate has a required coverage check after the pack's test checks: new-code coverage against the base branch (--base <ref>) must be at least 70% (WORKFLOW §6). gate.coverage_min may only raise that; gate.coverage_exclude adds exclusions to the pack's coverage.exclude (Go: cmd/**, internal/compose/**, testdata) and generated files. Only the pack's source_globs minus its test_globs count. In flag packs the race-guard hook denies a test command without the race flag. Coverage prerequisites (a coverage plugin or config) are set up by the <id>-engineer in wave 0; the language skill names them.
Reading the coverage table: each row is a changed package, module or file with its new-code percentage. The check fails when a row is under 70%, when a changed file's new lines are all untested, or (Go) when a changed package has no tests. Fix by adding tests for the listed uncovered file:line ranges, never by excluding code or lowering the bar. Scoped run for one agent: loops gate --only <the pack's test checks>,coverage --pkgs <owned> (Go: --only test,coverage). Coverage is still evidence for claims, not a goal in itself: assert behaviour.
Per language
The test runner, coverage format, race mode and substitute of each pack live in specs/languages/<id>/pack.json (loops lang show <id>) and skills/lang-<id>/testing.md. loops lang list shows every pack's race mode and coverage format. A language with no pack has no gate: say so in the claims report.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

