Sweep — Testing & QA agent for Claude Code
Cheap mechanical sweeps - find every place something is used across the repos, run test suites or linters and report only the failures, digest a long log into what matters.
How to install Sweep
Installs to ~/.claude/agents/chester0-claude-kit-sweep.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/chester0/claude-kit/HEAD/agents/sweep.md -o ~/.claude/agents/chester0-claude-kit-sweep.md Restart Claude Code, or start a new session, for it to be picked up.
What Sweep does
name: sweep description: Cheap mechanical sweeps - find every place something is used across the repos, run test suites or linters and report only the failures, digest a long log into what matters. Changes nothing. model: haiku effort: low omitClaudeMd: true tools: Read, Grep, Glob, Bash
You do one mechanical job fast and report only what matters. A long context costs more per token: the context rule in the house rules below matters most for you.
- Searching: list every match as file:l
Alternatives in Testing & QA
- Claude Code Test Agent — Tests all 27 Claude Code hooks by logging each event to tests-agents-hook/agent-hook-fired.log 321 ★
- Backend Test Architect — Use this agent when you need to design, implement, or review unit tests for NextJS backend components followin 76 ★
- Swift Debugger — Root-cause analysis for Swift/iOS failures — compiler errors, test failures, crashes, data races, SwiftUI view 33 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Prober
Runs a build, test, lint, or any command and returns a pass/fail probe (exit code, warn/error counts) instead
Test Tier Sweep
Periodic sweep across all 12 Zanix repos for a test file placed in the wrong @tests/ tier (unit/integration/fu
Log Digger
Triage high-volume output — cloud logs, container logs, test suite runs, build failures, long SQL result sets
Worker Sonnet
Tier-2 worker on Sonnet. Use for well-specified, bounded work - implementing a decided design, writing tests,
Runner
Test-suite execution on Opus 5. Use to run pytest, vitest, or the e2e suite and get back a digest of failures
QA Engineer Marvin
Use to validate that code matches requirements, or for adversarial code-quality review (execution-focused: run
Related Skills
Cluster Triage Skill
Ordered read-only diagnostic sweep of the MacMountain EKS clusters (customer360-prod, macmtn-dev-cluster) when
Rename Code
Safely rename a symbol/string across the codebase — grep all refs, scope-check, apply, verify with build + zer
No More Fomo
Never miss what matters in AI. Daily digest from 16 Twitter KOLs, 3 AI labs, 4 podcasts, arxiv topic search, H