
交件前,找出漏檔、錯版與審核回條不一致。需要檔案與已接受的清單;Python 本機執行,不需模型。Catch missing files, wrong versions and mismatched review receipts. Local Python download; accepted manifest required. No model needed.

Systematic QA of CLI tools and pipelines across multiple real workspaces using SFDPOT heuristics. Executes structured test flows, evaluates consistency oracles, collects evidence, and produces a structured bug report with P1/P2/P3 priorities and a release readiness verdict.

I'm a full-stack engineer who's shipped and maintained Next.js (App Router), React, and TypeScript apps under real production load. I kept watching AI write code that looked done - then leaked NEXT_PUBLIC_ secrets, swallowed errors, and skipped every state but "success." Shipwright distills the senior-engineer checklist I reach for when things break at 2 a.m. - secrets, auth, input validation, caching, and the unhappy paths - so your AI ships production-grade code, not demos.

在发布前检查 README、论文或报告中的声明是否仍与当前 CSV/JSON 结果一致,输出带证据的 SUPPORTED、STALE 或 UNSUPPORTED 闸门报告。

Find the inventory-sync bugs that don't throw errors — before your accountant does. Hand over 3 CSV exports (ERP stock, SKU mapping, channel catalog) and get back the exact SKUs silently mis-syncing right now, ranked by severity and explained. No API keys, no credentials — read-only by design.

From a local ISSUE file to a verified, committed deliverable — with a fresh-subagent acceptance gate that catches wrong screenshots and leaked logic.