本文へ移動
cccskills
無料GitHub で公開

package-update-check-and-plan

Read-only workflow for discovering uploaded deployment artifacts, comparing them with current deployment state, and preparing a safe update and rollback plan.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md4.1 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Package Update Check and Plan

Purpose

Identify whether new deployment artifacts exist, compare them with the currently deployed components, and produce a safe update plan with backup, verification, and rollback steps. The default workflow stops before any deployment or file modification.

When to use

  • A user says a new package, build artifact, archive, JAR/WAR, image, or frontend bundle was uploaded.
  • A user asks whether a project can be updated.
  • Multiple components need coordinated update planning.
  • You need a deployment plan before touching production files.

Safety rules

  • Default workflow is read-only.
  • Do not copy, unpack, replace, delete, chmod, chown, restart, reload, migrate, or clean files without explicit approval.
  • Do not reveal real upload paths, production paths, usernames, domains, IPs, database strings, or service names in public output.
  • Replace identifiers with <PROJECT_NAME>, <PROJECT_ROOT>, <UPLOAD_DIR>, <ARTIFACT_NAME>, <SERVICE_NAME>, and <APP_DOMAIN>.
  • Never trust a partially uploaded artifact. Size and checksum comparison belong in the plan before replacement.
  • Separate current-state verification from deployment action.

Read-only workflow

  1. Clarify the deployment target and artifact source location using placeholders if preparing a public report.
  2. List candidate artifacts in the upload/staging area by type, size, and timestamp. Do not unpack them.
  3. Identify currently deployed components: backend artifact, frontend bundle, service unit, reverse proxy config, config directory, data directory, and logs directory.
  4. Compare current vs candidate metadata: size, timestamp, version string if available, checksum if safe to compute, and component mapping.
  5. Check current service and port state without changing anything.
  6. Check whether the target machine differs from the current machine. If remote deployment is involved, plan source and target verification separately.
  7. Identify configuration compatibility risks: port changes, required environment variables, database migration expectations, reverse proxy upstreams, static asset layout, file ownership needs.
  8. Draft an update sequence with explicit approval gates: backup, artifact verification, frontend update, backend update, service restart, health checks, rollback.
  9. Draft rollback criteria and rollback steps, but do not execute them.

Evidence to collect

  • Artifact list with redacted names, size, timestamp, and type.
  • Current deployed component inventory.
  • Version/checksum comparison if available.
  • Current service state and listening port summary.
  • Reverse proxy/upstream mapping summary without real domains or IPs.
  • Config compatibility concerns.
  • Backup targets and rollback prerequisites.
  • Remote/local machine distinction.

Stop conditions

  • Artifact upload is incomplete or checksum cannot be verified.
  • Target host is ambiguous.
  • The update requires schema migration, data deletion, or destructive replacement and no approval exists.
  • Current state cannot be safely inventoried.
  • The package may contain secrets or private customer data.

Optional user-approved actions

Only after explicit user approval:

  • Create backups.
  • Transfer artifacts to a target host.
  • Verify checksums on source and target.
  • Unpack archives into staging.
  • Replace deployed files.
  • Reload or restart services.
  • Run post-deployment health checks that may touch write paths.
  • Execute rollback.

Final report format

Summary: <whether update appears possible>
Project: <PROJECT_NAME>
Candidate artifacts:
- <ARTIFACT_NAME>: <type/size/timestamp/checksum status>
Current deployment:
- Backend: <summary>
- Frontend: <summary>
- Service: <summary>
- Proxy: <summary>
Compatibility risks: <bullets>
Proposed update plan: <numbered approval-gated steps>
Rollback plan: <numbered approval-gated steps>
Read-only conclusion: <ready / blocked / needs more info>
Needs approval: <specific actions>

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Route CEO-first multi-window work into owner-led role loops. Use for 总控/CEO、架构/CTO、内容主编、角色派发与复用、模型路由、来源窗口回调、role-windows 台账、skill 命中统计或 multi-window loop engineering。

日本語の概要は準備中です。原文の説明を表示しています。

Dirtytrii/codex-skills112026年10月9日 更新

Read-only first workflow for diagnosing application incidents by collecting service, log, network, configuration, and functional evidence before proposing any fix.

日本語の概要は準備中です。原文の説明を表示しています。

Dirtytrii/codex-skills112026年10月9日 更新

Run user-authorized black-box web security tests and evidence-based reports. Use for 黑盒安全测试、渗透测试、数据泄露、未授权访问、公开 JS/API、登录爆破防护、CORS 或安全响应头检查。

日本語の概要は準備中です。原文の説明を表示しています。

Dirtytrii/codex-skills112026年10月9日 更新

Choose the narrowest reliable browser surface across Codex Browser, signed-in Chrome, Playwright CLI, and platform scripts. Use for browser automation, logged-in workflows, UI inspection, screenshots, form filling, publish previews, or frontend QA.

日本語の概要は準備中です。原文の説明を表示しています。

Dirtytrii/codex-skills112026年10月9日 更新

提议并执行 rubric 或 bucket 升级。两种模式:**完整 rubric bump**(最高风险动作,5 步强制 + 跨模型审核)和 **--bucket-only 轻量重校**(只换 bucket 边界,不动 rubric 公式)。**Phase 2 强制走 cheat-score-blind sub-agent 给校准池重打分**——不接受 self-scored fallback。触发词:"升级 rubric"/"bump rubric"/"更新公式"/"我想加一个维度"/"调整权重"/"重校桶"/"recalibrate bucket"。

日本語の概要は準備中です。原文の説明を表示しています。

Dirtytrii/codex-skills112026年10月9日 更新

初始化 cheat-on-content 内容校准工作区。仅在用户明确开始该工作流,或其子 skill 确认缺少 .cheat-state.json 时使用;不响应一般项目初始化、安装诊断或 Codex 首次会话。

日本語の概要は準備中です。原文の説明を表示しています。

Dirtytrii/codex-skills112026年10月9日 更新

Dirtytrii のスキルをすべて見る

このスキルの問題を報告する