From retort
This skill should be used when the user asks to "run tests", "generate tests", "check coverage", "write tests for X", "what tests are missing", "triage test failures", "fix failing tests", or mentions xUnit, Vitest, cargo test, dotnet test. Provides the full testing workflow across .NET, TypeScript, and Rust stacks.
How this skill is triggered — by the user, by Claude, or both
Slash command
/retort:testing-agentThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Cross-stack testing workflow for .NET (xUnit), TypeScript (Vitest), and Rust (cargo test). Covers test generation, execution, triage, and coverage analysis.
Cross-stack testing workflow for .NET (xUnit), TypeScript (Vitest), and Rust (cargo test). Covers test generation, execution, triage, and coverage analysis.
Two agents handle the testing workflow:
Invoke them explicitly (use the test-generator agent) or they activate on matching phrases.
Detect the right test command from project signals:
| Signal | Command |
|---|---|
*.csproj with xUnit refs | dotnet test |
vitest in package.json | pnpm test or pnpm vitest run |
Cargo.toml in crate root | cargo test |
Always run from the appropriate root (solution root for .NET, package root for TS, crate root for Rust).
Context determines scope — a single service change warrants scoped; a shared library change warrants full.
When tests fail:
Run coverage commands from references/mystira-stacks.md for stack-specific flags.
Minimum workflow before a PR:
git diff main --name-only)At session start in a new project, coverage-guard will ask for a coverage threshold if one isn't in .claude/retort.local.md. Answer once — it won't ask again that session.
Create .claude/retort.local.md in any repo to configure per-project behaviour:
coverage_threshold: 80
test_style: unit-only
coverage_check: on_demand
baseline_branch: main
references/mystira-stacks.md — Mystira-specific test commands, paths, xUnit/Vitest/cargo conventionsGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub phoenixvc/retort --plugin retort