{
  "markdown": "# `sddw-gemini`\n\nSpec-Driven Development Workflow for [Gemini CLI](https://github.com/google-gemini/gemini-cli).\n\n- Write **requirements**, optionally **analyse the codebase**, then **design** architecture, then **taskify** (as hybrid task files referencing `design.md`), then **implement** each task separately, **task-review** each implemented task, then **verify** the result, then **self-improve** the workflow\n- The agent guides you through every step — researches, proposes options, confirms your decisions\n- Every step produces exactly one spec type. Every step reads specs from previous steps.\n- Two interaction modes: guided dialog (default) or fully `--auto`\n- Lightweight and easily customizable — just markdown and TOML files, no runtime dependencies\n\nAlso available for [Claude Code](https://github.com/sermakarevich/sddw).\n\n## Why\n\nThe standard way to use AI coding agents is short, interactive prompts: describe what you want, get code, fix it, repeat. This works for small tasks but breaks down for anything non-trivial — context gets lost between sessions, architectural decisions live only in chat history, and there's no artifact a teammate can review before code is written.\n\n`sddw` inverts this. Instead of prompting for code, you collaborate with the agent to write **specifications** — requirements, architecture, interface contracts, task breakdowns. The specs become the primary artifact: reviewable by peers, version-controlled, persistent across sessions. Code generation is then a mechanical step guided by approved specs, not a creative leap from a vague prompt.\n\nDetailed specifications reduce AI code errors by up to 50% (Piskala, 2026), security defects by 73% (Marri, 2026), and architecture-misaligned PRs by 60% (GitHub Spec Kit). `sddw` is designed for medium to large projects that don't fit into a single context window. By splitting work into discrete steps — requirements, codebase analysis, design, per-task implementation — each step operates within a focused context where models are more accurate, rather than a sprawling conversation where critical details get lost.\n\n## Install\n\n```bash\ngemini extensions install https://github.com/christianmerkwirth/sddw-gemini\n```\n\nFor development (link from local repo):\n\n```bash\ngit clone https://github.com/christianmerkwirth/sddw-gemini\ncd sddw-gemini\ngemini extensions link .\n```\n\n## Commands\n\n| Command | Description |\n|---------|-------------|\n| `/sddw:requirements <feature> [--auto]` | Generate requirements spec |\n| `/sddw:code-analysis <feature> [--auto]` | Analyse existing codebase (optional) |\n| `/sddw:design <feature> [--auto]` | Generate cross-cutting `design.md` |\n| `/sddw:taskify <feature> [--auto]` | Generate hybrid task files from `design.md` |\n| `/sddw:design_and_taskify <feature> [--auto]` | Combined alias: design + taskify in one shot |\n| `/sddw:implement <feature> --task <N> [--auto]` | Implement a single task |\n| `/sddw:task-review <feature> --task <N> [--auto]` | Review a single implemented task |\n| `/sddw:verify <feature> [--auto]` | Verify implementation against requirements |\n| `/sddw:self-improve <feature> [--auto]` | Analyse execution and improve workflow |\n| `/sddw:chat <feature> [--auto]` | Fast-track interaction with an existing feature |\n| `/sddw:help [list \\| status <feature>]` | Workflow overview and feature status |\n\n## Interaction Modes\n\nEvery step supports two interaction modes:\n\n| Mode | Flag | Behavior |\n|------|------|----------|\n| **Interactive** | *(default)* | Full guided dialog — one question at a time, every section confirmed |\n| **Auto** | `--auto` | Fully autonomous — no questions, best-judgment decisions |\n\n## Workflow\n\n### 1. Requirements\n\n```\n/sddw:requirements <feature-name> [--auto]\n```\n\nCollaboratively produce a requirements spec through guided dialog:\n\n- **Discover** — understand the feature through one-at-a-time questions\n- **Research & Propose** — research SOTA, codebase, domain; propose each section with ranked options\n- **Confirm & Generate** — user approves each block, spec is written\n\nOutput: `.sddw/<feature-name>/requirements.md`\nSections: Purpose, User Stories, Functional Requirements, Acceptance Criteria, Constraints\n\n### 2. Code Analysis (optional)\n\n```\n/sddw:code-analysis <feature-name> [--auto]\n```\n\nAnalyse the existing codebase to ground design decisions in reality:\n\n- **Discover** — understand which areas of the codebase matter most\n- **Research & Propose** — scan for patterns, interfaces, flows, conventions\n- **Confirm & Generate** — user approves each section, analysis is written\n\nOutput: `.sddw/code-analysis.md` (shared across features)\n\nSkip this step for greenfield projects with no existing codebase.\n\n### 3. Design\n\n```\n/sddw:design <feature-name> [--auto]\n```\n\nProduce a cross-cutting `design.md` through guided dialog:\n\n- **Discover** — understand architectural preferences and constraints\n- **Research & Propose** — propose architecture, data models, contracts, and decisions\n- **Confirm & Generate** — user approves each block, `design.md` is written\n\nOutput: `.sddw/<feature-name>/design/design.md`\n\n### 4. Taskify\n\n```\n/sddw:taskify <feature-name> [--auto]\n```\n\nBreak the feature into hybrid task files based on requirements and design:\n\n- **Discover** — understand task granularity preferences\n- **Research & Propose** — propose task breakdown in dependency order\n- **Confirm & Generate** — user approves task list, task files are written\n\nOutput:\n\n```\n.sddw/<feature-name>/\n└── design/\n    ├── design.md\n    └── tasks/\n        ├── task-1-<slug>.md  # hybrid: files, criteria, references design.md\n        ├── task-2-<slug>.md\n        └── ...\n```\n\nEach task file includes task-specific details inline and references `design.md` for architecture, models, and shared contracts, so the implementation agent has full context without duplication.\n\n### Design & Taskify (Combined Alias)\n\n```\n/sddw:design_and_taskify <feature-name> [--auto]\n```\n\nRuns the design and taskify steps in one shot. Use this for small features where iterating on architecture independently is not needed.\n\n### 5. Implement\n\n```\n/sddw:implement <feature-name> --task <N> [--auto]\n```\n\nExecute a single task from the design spec:\n\n- **Discover** — select task, check dependencies, gather context\n- **Research & Propose** — scan codebase, propose implementation approach and TDD applicability\n- **Execute** — implement following TDD protocol, commit protocol, and deviation handling\n\nAfter each task, a completion report (`task-N-<slug>.done.md`) is written to `implement/tasks/`.\n\n### 6. Task Review\n\n```\n/sddw:task-review <feature-name> --task <N> [--auto]\n```\n\nA focused, task-level quality gate that runs after each `implement` and before feature-level `verify`. It reviews one task's diff — not a rewrite — and routes back to `implement` when changes are needed:\n\n- **Assess** — load the task file, completion report, and design; identify the diff from the task's commits\n- **Review** — run the task's tests and review the changes across six dimensions: criteria conformance, design conformance, convention conformance, code quality, test quality, and deviation integrity\n- **Report & Decide** — classify findings (Blocker / Major / Minor) and issue a verdict: **APPROVED** or **CHANGES REQUESTED**\n\nOutput: `.sddw/<feature-name>/task-review/task-<N>-<slug>.review.md`\n\nSteps 5 and 6 form a per-task loop: implement a task, review it, then move to the next task. When all tasks are implemented and approved, proceed to verify.\n\n### 7. Verify\n\n```\n/sddw:verify <feature-name> [--auto]\n```\n\nVerify the implementation against requirements after all tasks are complete:\n\n- **Assess** — load artifacts, detect test runner, check task completion status\n- **Verify** — run test suite, cross-check each FR's acceptance criteria, review done criteria\n- **Report & Remediate** — produce verification report, create remediation tasks if issues found\n\nOutput:\n\n```\n.sddw/<feature-name>/\n└── verify/\n    └── report.md    # FR-by-FR pass/fail, test results, deviations, warnings\n```\n\nIf issues are found, remediation tasks are created as additional task files in `design/tasks/`. These can be executed with `/sddw:implement` and then verified again.\n\n### 8. Self-Improve\n\n```\n/sddw:self-improve <feature-name> [--auto]\n```\n\nAnalyse the completed feature's execution across all workflow steps. Identify what went wrong (or could be better) and propose concrete improvements to the workflow itself:\n\n- **Analyse** — extract signals: deviations, difficulties, remediation task origins, spec gaps\n- **Diagnose** — classify findings by workflow step, identify patterns, propose improvements\n- **Apply** — present proposals with diff previews, apply approved changes to workflow files\n\nOutput:\n\n```\n.sddw/<feature-name>/\n└── self-improve/\n    └── report.md    # findings, proposals, applied/skipped changes\n```\n\n### Chat\n\n```\n/sddw:chat <feature-name> [--auto]\n```\n\nFast-track interaction with a feature that already has artifacts. Skips the full questionnaire ceremony — load context and talk.\n\n- **Questions** — ask anything about the feature\n- **Spec updates** — edit requirements, FRs, acceptance criteria, or task files in-place\n- **Quick implementation** — small code changes following TDD and commit protocols\n- **Status** — check feature progress\n\n### Help\n\n```\n/sddw:help [list | status <feature-name>]\n```\n\n- `/sddw:help` — workflow overview\n- `/sddw:help list` — list all features with progress indicators\n- `/sddw:help status <feature-name>` — detailed feature status\n\n## Output Structure\n\nAll artifacts live under `.sddw/` in the project root:\n\n```\n.sddw/\n  <feature>/\n    requirements.md           # Step 1 output\n    design/\n      design.md               # Cross-cutting design artefact\n      tasks/\n        task-1-<slug>.md      # One file per implementation task\n        task-2-<slug>.md\n    implement/\n      tasks/\n        task-1-<slug>.done.md # Completion reports\n    task-review/\n      task-1-<slug>.review.md # Per-task review reports (verdict + findings)\n    verify/\n      report.md               # FR-by-FR pass/fail, test results\n    self-improve/\n      report.md               # Findings, proposals, applied changes\n  code-analysis.md            # Optional shared codebase analysis\n```\n\n## License\n\nMIT\n",
  "bytes": 10288,
  "sha": "5625648e8c8c1b16034223d1cabbaabaeda7ef62fa6fd0c84dc5ebca1a864e6f",
  "repo_slug": "christianmerkwirth/sddw-gemini",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_christianmerkwirth_sddw_gemini_f5dd383d/readme"
}