技能 engineering-loop

📦

engineering-loop

Name: engineering-loop
Author: phelan164

內容修訂 r1 安全 ⚙️ 外部命令

Complete Repository Changes with Verified Evidence

Repository changes can fail when implementation, testing, and review are disconnected. This skill manages a complete local engineering loop with explicit evidence and stop conditions.

支援: Claude Codex Code(CC)

📊 75 尚可

Maintain this Skill? Claim it 📊 75 尚可

使用我的智慧代理安裝

將此請求複製給你的智慧代理，其中包含權威 Skill 頁面和清單。

智能體請求

Review the Skillstore skill "engineering-loop" from https://skillstore.io/skills/phelan164-engineering-loop.md and its manifest at https://skillstore.io/api/skills/phelan164-engineering-loop/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

你的智慧代理仍應顯示其計畫，並依安全政策請求必要確認。

Agent 可讀資源

當 AI Agent、爬蟲或腳本需要乾淨脈絡、而不是讀取完整頁面時，請使用這些連結。

Markdown 詳情 GET /skills/phelan164-engineering-loop.md 已簽名 manifest GET /api/skills/phelan164-engineering-loop/manifest 已簽名 lockfile GET /api/skills/phelan164-engineering-loop/lockfile

測試它

正在使用「engineering-loop」。 Fix the duplicate invoice submission defect and add regression coverage.

預期結果:

Fixed duplicate submission handling in the owning request path.
Added a regression test covering repeated invoice requests.
Focused tests and the required broader checks passed.
Reviewed the final diff; no unresolved consequential findings remain.

正在使用「engineering-loop」。 Add account export with authorization checks and report any blockers.

預期結果:

Implemented authorized account export with explicit ownership validation.
Added success, forbidden, and malformed-request coverage.
Focused validation passed, but the full integration suite requires unavailable infrastructure.
The remaining blocker and exact required service were documented.

正在使用「engineering-loop」。 Investigate an intermittent cache invalidation failure.

預期結果:

Created a bounded reproducer and recorded the distinguishing failure signature.
Tested ranked timing and key-normalization hypotheses independently.
Isolated the cause, added regression coverage, and removed temporary instrumentation.
The original reproduction and affected repository checks passed.

安全審計

安全

v1 • 2026/7/31 開啟版本化報告

The two static findings are false positives caused by prose and inline Markdown formatting. The skill defines a bounded local engineering workflow without executable shell logic, prompt injection, or data-exfiltration intent.

已掃描檔案

212

分析行數

複核項目

已忽略的誤報

風險因素

⚙️ 外部命令 (1)

SKILL.md:13

最近一次完成的靜態與語意稽核未偵測到已確認的安全發現。這並不能證明該 Skill 沒有副作用。

審計單位: codex

分享與引用此報告

分享版本化評估報告、中性徽章、嵌入卡片與引用。Skillstore 只報告證據，不判斷該 Skill 是否安全。

開啟版本化報告

複製報告連結

https://skillstore.io/skills/phelan164-engineering-loop/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/phelan164-engineering-loop/security.svg)](https://skillstore.io/skills/phelan164-engineering-loop?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/phelan164-engineering-loop?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/phelan164-engineering-loop/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/phelan164-engineering-loop.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

學術引用 (APA · BibTeX · CFF)

APA 引用

phelan164. (2026). engineering-loop security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/phelan164-engineering-loop/audits/1

BibTeX 引用

@techreport{phelan164-phelan164-engineering-loop-2026,
  author = {phelan164},
  title = {engineering-loop security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/phelan164-engineering-loop/audits/1},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "engineering-loop security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "phelan164"
date-released: "2026-07-31"
url: "https://skillstore.io/skills/phelan164-engineering-loop/audits/1"
identifiers:
  - type: other
    value: "skillstore:phelan164-engineering-loop:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 評分

評分依據證據置信度: 中

架構

可維護性

內容

社群

規範符合度

你可以打造什麼

Implement a Focused Feature

Complete a repository feature from baseline inspection through tests, diff review, and final evidence.

Resolve a Reproducible Defect

Reproduce a defect, add regression coverage, implement a bounded fix, and verify affected behavior.

Coordinate a Cross-Stack Change

Track acceptance criteria, validation levels, contract effects, and blockers across several repository layers.

試試這些提示

Fix a Small Defect

Use the engineering loop to reproduce and fix [defect]. Preserve unrelated changes, run focused tests, review the diff, and report evidence.

Build a Feature

Implement [feature] with these acceptance criteria: [criteria]. Establish the baseline, add appropriate tests, run required checks, and summarize residual risks.

Investigate an Intermittent Failure

Investigate [failure] using a bounded reproducer and ranked hypotheses. Isolate the cause, add regression coverage, implement the smallest fix, and reverify.

Execute a High-Risk Cross-Stack Change

Complete [change] across [layers]. Track contracts, migrations, authorization boundaries, data integrity, validation levels, review findings, and explicit stop conditions.

最佳實務

Define observable acceptance criteria and a safe baseline before changing code.
Run the narrowest relevant check after each bounded change, then complete required broader validation.
Preserve unrelated work and report unrun checks, residual risks, and blockers explicitly.

避免

Do not weaken assertions, remove coverage, or hide errors to make validation pass.
Do not repeat identical failing commands without classifying the failure and revisiting assumptions.
Do not deploy, migrate, merge, publish, or query production without explicit authorization.