St Work — Testing skill for Claude Code
Execute a change end to end: frame, research, plan, build with sub-agents, prove with gates, review loop, QA checkpoint, proof block.
How to install St Work
Installs to ~/.claude/commands/zomarit-stamity-st-work.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/zomarit/stamity/HEAD/.claude/commands/st-work.md -o ~/.claude/commands/zomarit-stamity-st-work.md Restart Claude Code, or start a new session, for it to be picked up.
What St Work does
description: "Execute a change end to end: frame, research, plan, build with sub-agents, prove with gates, review loop, QA checkpoint, proof block."
/st-work
Execute one change end to end. Five phases — Frame, Understand, Plan, Build, Prove — closing with a QA human checkpoint and a machine-readable proof block.
Phase 0 — Frame
Seconds, not ceremony. In order:
- Parse intent. Restate the request as one outcome sentence plus in-scope / out-of-scope bullets. Ambiguity gat
Alternatives in Testing
- Deslop Reborn — One iteration of the IronClaw Reborn de-slop loop — take ONE Reborn crate, fan out parallel review sub-agents 12.6k ★
- Route Research For Testing — route research for testing skill 9.5k ★
- Flow Spec — NLSpec authoring — use when you need a structured specification from multi-AI research and consensus 2.8k ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Build Module
Build a module from a spec with integrated quality gates and mid-build checkpoint reviews. Generates code task
Agent Proof Codebase
Skill for Claude Code and opencode: deterministic quality gates (invariant specs, mutation testing, fail-to-pa
QA Phase
Run QAPhase — the autonomous acceptance-and-ship QA gauntlet. Proves a finished story end-to-end (spec-gap/gri
Claude Code Hooks Cookbook
Copy-paste Claude Code hooks that actually do something — auto-format, block dangerous commands, test gates, n
Test Review
Write-then-verify test review — write tests in your normal flow, then mechanical gates (red-green + mutation)
Cmd Verify
Verify phase (PROVE) format, lint, type-check, scan, test, and githook gate with a fix/review loop. Use to lea
Related Agents
Gantry Verifier
Runs the readiness gate. Use to execute the project's lint/test/build from its own .claude/gates.sh, the check
Agent Workflow - Automated Development Pipeline
Execute complete development workflow using intelligent sub-agent chaining with quality gates.
Research Loop
Autonomous iterative research loop — hypothesis, search, ingest, evaluate, keep/discard via checkpoint. Max 3