Agent Plugins Marketplace
All plugins

dsh-codex-workflow

v0.100.3

A ten-skill development workflow for AI software projects, distributed as a Codex plugin for product confirmation, planning, implementation, verification, repair, review, delegation, research, and release.

Codex10 Skills

By DSH AI Software WorkflowLicense: Apache-2.00 GitHub starsUpdated 2 weeks ago

Directory evidence

Runtimes
Codex
Parsed components
10 skill or MCP entries
Source updated
Sep 9, 2026
Manifest status
Canonical path parsed

The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology

Install dsh-codex-workflow for Codex

Installs for the current user
codex plugin marketplace add IchenDEV/agent-plugin-mkt
codex plugin marketplace upgrade agent-plugin-marketplace
codex plugin add dsh-codex-workflow@agent-plugin-marketplace

Paste and run these commands in a terminal with Codex. They add and refresh the PluginsMP catalog, then install this plugin.

The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.

Get the source manually
git clone https://github.com/v20227/dsh-codex-workflow

Clone the source repository, then follow its setup instructions to add the plugin to a compatible client. The repository root is the plugin root.

Plugin files

dsh-codex-workflow/
├── .codex-plugin/plugin.json
├── skills/sakura-code/SKILL.md
├── skills/sakura-delegate/SKILL.md
├── skills/sakura-dev/SKILL.md
├── skills/sakura-fix/SKILL.md
├── skills/sakura-plan/SKILL.md
├── skills/sakura-release/SKILL.md
├── skills/sakura-research/SKILL.md
├── skills/sakura-review/SKILL.md
├── skills/sakura-spec/SKILL.md
└── skills/sakura-verify/SKILL.md

Included Skills10

sakura-codeskills/sakura-code/SKILL.md

Implement a confirmed software feature, repair, or refactor within its current scope and architecture. Use when the expected result and verification method are clear enough to change code; do not use for read-only diagnosis or review.

sakura-delegateskills/sakura-delegate/SKILL.md

Keep software-development context reliable by frequently offloading bounded exploration, research, test analysis, diagnosis, review, and isolated implementation to native subagents or separate Codex tasks, then closing or archiving them after verified handoff. Use throughout non-trivial projects, before raw evidence fills the main task, and whenever a repeated problem needs a clean context.

sakura-devskills/sakura-dev/SKILL.md

Run software development from the user's natural-language goals, generating and maintaining the working PRD, acceptance criteria, and development plan. Use to start or resume a project, coordinate workstreams and supporting tasks, decide what happens next, and carry an accepted milestone through implementation, verification, repair, review, and release.

sakura-fixskills/sakura-fix/SKILL.md

Diagnose and repair a reproducible software failure without repeating the same unsuccessful patch. Use after a test, build, runtime, integration, package, or review failure has produced concrete evidence.

sakura-planskills/sakura-plan/SKILL.md

Convert an accepted software outcome and current project facts into the smallest executable development plan. Use when work spans multiple meaningful changes, dependencies, integration points, or recovery checkpoints; do not use merely to restate a simple edit.

sakura-releaseskills/sakura-release/SKILL.md

Prepare and verify a user-ready software release from a known source state. Use when the user asks for a clean version, distributable archive, installer, release candidate, deployment, publication, or clarification of development versus released versions.

sakura-researchskills/sakura-research/SKILL.md

Verify external technical facts that materially affect an active software decision, including official platform behavior, framework compatibility, APIs, standards, open-source reference projects, and relevant implementation patterns. Use when the current project and available evidence cannot support a reliable decision or when external code may affect provenance and licensing.

sakura-reviewskills/sakura-review/SKILL.md

Review a software change against the accepted outcome, current code, and actual verification evidence, then recommend or record whether it passes, needs repair, or is blocked according to the current context role. Use before closing a workstream or when the user requests a code or completion review.

sakura-specskills/sakura-spec/SKILL.md

Convert the user's natural-language software goal, feature request, complaint, screenshot, or defect report into a maintained working PRD, bounded user-visible outcome, and executable acceptance criteria. Use throughout Sakura development so Codex owns this translation by default while accepting user edits.

sakura-verifyskills/sakura-verify/SKILL.md

Verify software changes through current executable evidence, including targeted tests, type checks, builds, browser behavior, APIs, CLIs, integration, and package checks. Use after implementation, after repair, or when a reported failure must be reproduced.

Plugin manifests1

.codex-plugin/plugin.json
{
  "name": "dsh-codex-workflow",
  "version": "0.100.3",
  "description": "A ten-skill development workflow for AI software projects, distributed as a Codex plugin for product confirmation, planning, implementation, verification, repair, review, delegation, research, and release.",
  "author": {
    "name": "DSH AI Software Workflow"
  },
  "license": "Apache-2.0",
  "keywords": [
    "ai software",
    "software development",
    "workflow",
    "codex",
    "verification",
    "release"
  ],
  "skills": "./skills/",
  "interface": {
    "displayName": "DSH AI 软件工作流",
    "shortDescription": "把任何 AI 软件项目推进到可验证发布",
    "longDescription": "面向任何 AI 软件项目,从自然语言目标开始,完成产品确认、任务拆解、实现、验证、连续纠错、审查和发布。当前以 Codex 插件形式分发,工作流本身不限制项目使用的模型、框架或 AI 工具。",
    "developerName": "DSH AI Software Workflow",
    "category": "Productivity",
    "capabilities": [
      "Interactive",
      "Read",
      "Write"
    ],
    "defaultPrompt": [
      "把这个软件想法确认清楚并拆成可执行任务",
      "检查当前结果,验证并连续修复直到可用",
      "把已完成的功能审查并整理成可发布版本"
    ]
  }
}

If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.

[dsh-codex-workflow on Agent Plugins Marketplace](https://pluginsmp.com/plugins/dsh-codex-workflow)