do-test

Runs builds, test suites, and coverage verification through a compiled contract across the repo's detected toolchain.

Updated Jul 9, 2026
One-click install
npx skills add https://github.com/khoavu882/do-flow --skill do-test-khoavu882
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: do-test
Source: https://github.com/khoavu882/do-flow/tree/main/core/shared/skills/do-test
Command: npx skills add https://github.com/khoavu882/do-flow --skill do-test-khoavu882

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Running tests and builds across heterogeneous repositories often means guessing which toolchain is present, manually invoking runners, and losing track of which checks actually ran. This Skill compiles a verification contract first, then executes and reports every tier against it so no check is silently skipped or falsely reported as passed. ## Core Features & Use Cases - Contract-first verification: Compiles the tier list, risk level, and detection gaps via the DoFlow runtime before any command runs, so the expected scope is stated up front. - Toolchain auto-detection: Detects the repo's existing build and test runners from its manifests instead of requiring manual configuration. - Honest reporting: Reports every tier status (PASS, FAIL, UNRESOLVED, SUBSUMED, NOT_APPLICABLE, SKIPPED, NOT_RUN) and marks runs INCONCLUSIVE when required tiers cannot be resolved. - Use Case: After implementing a feature, invoke the skill to verify the build compiles, the test suite passes, and coverage numbers are reported, then record the handoff so the next workflow stage knows the verification result. ## Quick Start Ask the assistant to run the test suite and verify the build for the current project using the do-test skill.

Frequently Asked Questions about do-test

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run tests and verify a build in a repo with an unknown toolchain?▼

Invoke the do-test skill, which detects the repo's build and test runners from its manifests and compiles a verification contract before running anything. It then executes each contract tier and reports pass, fail, or unresolved status for every one.

How do I check test coverage without configuring a coverage tool?▼

Coverage is reported from whatever the detected test runner already emits, including line and branch numbers when available. If the runner does not produce coverage data, the skill states that plainly rather than fabricating numbers.

What happens when a required test tier cannot be detected in the repo?▼

The tier is marked UNRESOLVED, and one unresolved required tier makes the entire run INCONCLUSIVE. The skill reports this verdict with the reason naming the missing tiers instead of narrowing the scope to force a green result.

Can I run tests in watch mode or with a clean build?▼

Yes, the --clean flag forces a clean compile before deterministic tiers run, and --watch launches interactive watcher mode when the detected runner supports it. Watch mode is an interactive session and produces no contract verdict.

What are the limitations of this test verification approach?▼

It will not write new test files, modify build configuration, or skip failing tests to force a pass. It also requires the DoFlow runtime installed via npx @khoavu882/doflow install, and it never reports a verdict the compiled contract did not produce.