Real Repo Testing — Testing skill for Claude Code
Use when testing buoy changes against real-world repos, validating drift detection accuracy, comparing before/after results, or preparing a release that needs real-repo validation.
How to install Real Repo Testing
Installs to ~/.claude/skills/ahoybuoy-buoy-real-repo-testing/SKILL.md
mkdir -p ~/.claude/skills/ahoybuoy-buoy-real-repo-testing && curl -fsSL https://raw.githubusercontent.com/ahoybuoy/buoy/HEAD/.claude/skills/real-repo-testing.md -o ~/.claude/skills/ahoybuoy-buoy-real-repo-testing/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Real Repo Testing does
name: real-repo-testing description: Use when testing buoy changes against real-world repos, validating drift detection accuracy, comparing before/after results, or preparing a release that needs real-repo validation.
Real-Repo Testing
This skill defines how to test buoy against real-world repos using buoy-lab.
Testing Tiers
| Tier | When | Speed | Command |
|---|---|---|---|
| Unit tests | Every change | ~30s | pnpm test (2512+ vitest tests) |
| Manual spot-ch |
Alternatives in Testing
- Fix Issue — by metabase - Addresses GitHub issues by taking issue number as parameter, analyzing context, implementing sol 46.5k ★
- Gstack — 15,000+ Garry Tan's exact Claude Code setup: 6 opinionated tools that serve as CEO, Eng Manager, Release Manag 33.9k ★
- Loki Spec Status — Check whether the spec has drifted from its lock using Loki's living-spec drift detection, and summarize the r 1.1k ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
QA Real World
Run free-form agent-based QA against real-world public repos. Each agent acts as a developer who just discover
Ship And Verify
Build the Mori desktop release, prompt user to install + run targeted tests, then read JSONL log to verify the
Think Stress Test Decision
Pressure-test a consequential decision before committing, by comparing options, surfacing the conditions that
Plan To Spec
Use when turning a planning note into a build spec, preparing a note for handoff to a coding-agent build, or c
Heal Playwright Tracer
Open-source statement-level Playwright tracer, purpose-built for AI agents. Analyzes test runs with increased
Decisionbench
One public test, any typed-decision model: accuracy, latency, cost, and accuracy when it is sure. Jev vs Claud
Related Agents
Statistical Specialist
Use for statistical analysis, hypothesis testing, and significance testing. Invoked for comparing tier results
Security Engineer Smythe
Use for security audits, auth/crypto/input validation reviews, dependency scanning, secret detection, or valid
Orbit Uat
Real-world testing in Docker WP environments. Spins up clean installs, runs flows, files bugs. Orchestrates fu