compare-test-runs — Claude Code skill: overview & setup
compare-test-runs compares two Tuist test executions to spot fresh failures, newly flaky suites, and execution duration spikes.
no license 265 installs ★ 35 updated 2026-03-27
npx skills add tuist/agent-skills --skill compare-test-runs -a claude-codeWhat compare-test-runs does
This skill evaluates test runs side by side. It accepts run identifiers, web dashboard URLs, or git branch references. When omitted, it defaults to checking your active branch against the default branch, which is typically main.
The logic relies entirely on Tuist CLI commands. The agent queries run data with tuist test show and itemizes individual cases via tuist test case run list using JSON output. It maps cases across modules, sorting them into clear categories: newly broken tests, resolved cases, newly flaky behavior, and runtime spikes exceeding 50%.
Without the tuist binary available in your shell environment, the skill cannot fetch test telemetry. Its finished summary highlights failing assertions alongside precise source paths and line numbers.
What it helps with
- detect newly broken tests between base and feature branches
- catch flaky test cases before merging pull requests
- isolate test duration regressions across modules
- extract failure lines and stack traces from test reports
Example prompts for the agent
- Compare test runs between main and my current branch using Tuist
- Inspect test run IDs run-123 and run-456 to show what broke
- Check the latest test results for branch feature-cache and list newly flaky tests
How to install compare-test-runs
1Skills CLI (Vercel)
npx skills add tuist/agent-skills --skill compare-test-runs -a claude-code2Manually: put the files in place
~/.claude/skills/compare-test-runs/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/compare-test-runs/
1Skills CLI (Vercel)
npx skills add tuist/agent-skills --skill compare-test-runs -a codex2Manually: put the files in place
.agents/skills/compare-test-runs/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add tuist/agent-skills --skill compare-test-runs -a cursor2Manually: put the files in place
.cursor/skills/compare-test-runs/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install tuist/agent-skills skills/compare-test-runs2Manually: put the files in place
.github/skills/compare-test-runs/SKILL.md1Skills CLI (Vercel)
npx skills add tuist/agent-skills --skill compare-test-runs -a gemini-cli2Platform CLI
gemini skills install https://github.com/tuist/agent-skills --path skills/compare-test-runsManually: put the files in place
.devin/skills/compare-test-runs/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add tuist/agent-skills --skill compare-test-runs -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about compare-test-runs
How to install compare-test-runs in Claude Code?
Run npx skills add tuist/agent-skills --skill compare-test-runs -a claude-code in your terminal.
What does the compare-test-runs skill do?
It queries Tuist test metadata to contrast two test runs, classifying failures, flaky tests, and execution time changes.
What prerequisite does this skill require?
It requires the tuist CLI installed and accessible, as the assistant invokes tuist test commands under the hood.