clean-architecture-rebuild
Same behavior; new seams / folders / decoupling.
Skill by nhouseholder
Chalk Labs
Need help deploying clean-architecture-rebuild?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/nhouseholder-clean-architecture-rebuild
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/clean-architecture-rebuild
What's inside
- Skill name
clean-architecture-rebuild- Description
- Same behavior; new seams / folders / decoupling. Convert code to clean architecture — separate concerns, increase modularity, reduce coupling.
- License
- MIT
- Instructions
- 80 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Clean architecture rebuild Behavior unchanged. New seams, folders, and decoupling so policy does not depend on I/O. `$ARGUMENTS` is the feature / path to rebuild, and/or `--report-only`. `--report-only` proposes the folder plan and does not move code. ## Hard gates - **Same behavior.** Inputs, outputs, errors, ordering, and side effects stay. If a test must change, the rebuild leaked a behavior change — stop. - One writer per shared key / file remains one writer. Do not split a store and leave two combiners. - Two sources for one fact stay a bug; the rebuild collapses them or keeps a single SoT with a named adapter. - Do not invent layers the repo already has under another name. Match neighboring module shape. - Do not add a library that duplicates one already in the repo. ## Vocabulary Use these words in the plan and the code: - **Policy** — the decision the product cares about - **Seam** — the interface you can test through - **Adapter** — I/O, HTTP, disk, UI, clock behind the seam - **Inward** — adapters depend on policy; policy never imports adapters One adapter is a hypothetical seam. Two adapters (or a test double plus production) make it real. ## 1. Scope the slice Named feature / path → that slice. No path → ask which job is being decoupled; do not rebuild the whole repo. Read existing docs (`AGENTS.md`, `CLAUDE.md`, `CONTEXT.md`, ADRs) so you do not reopen a decided seam. ## 2. Current coupling map For the slice, list: - Policy mixed into I/O (SQL in a route, fetch in a calculator, disk in a scorer) - Folder layout vs actual dependencies (`grep`/imports, not the folder names) - Tests that boot the world because there is no seam - Sacred choke points already documented (do not bypass them) ## 3. Folder / seam plan Propose the smallest plan that makes policy inward-only. Example shape — **adapt names to this repo**, do not paste a generic Clean Architecture tree into a JS worker or a Python pipeline that already has a house style: ``` <slice>/ <policy> # pure decisions, types, invariants <seam> # ports the policy calls <adapter> # HTTP, R2, SQLite, UI, clock <tests> # policy tests through the seam ``` For each move: from-path, to-path, what the seam is, what stays put. `--report-only` prints this plan and stops. No files moved. ## 4. Move (not `--report-only`) Order: 1. Characterization tests around the current behavior if none exist. 2. Extract the seam (interface / function boundary) **in place**. 3. Move policy to the new folder; leave adapters calling in. 4. Point tests at the seam. Run them after each move. 5. Delete the hollow old file. Do not leave a re-export maze unless a public import path must stay stable — then one shim, listed in the report. No feature work in the same change. No “while we’re here” refactors outside the plan. ## 5. Verify - Same tests green without rewriting assertions to match new behavior. - Policy modules do not import adapters (spot-check imports). - Public callers still compile / run. - One-writer and one-SoT invariants still hold. ## Done Plan executed or (report-only) proposed. Behavior unchanged. New seam is the test surface. Extra folders that do no decoupling were not added.
From nhouseholder/cursor-laptop-skills, MIT.
More from nhouseholder/cursor-laptop-skills
- 0.524score
debate
nhouseholder
Runs a formal 3-round debate between Claude (Proposer) and Gemini 3.1 Pro (Challenger), ending with an AI verdict via backend API.
SkillBFreshness: BClaude Code +9 - 0.524score
estimate
nhouseholder
Effort, time, and complexity estimation before starting a task.
SkillBFreshness: BClaude Code +9 - 0.524score
reflect
nhouseholder
Session retrospective — review what was accomplished in this session versus the original goal.
SkillBFreshness: BClaude Code +9 - 0.524score
whats-next
nhouseholder
Strategic next-action prioritizer: anchors on the project's north star, scores impact vs effort vs confidence vs decay, filters already-killed work, and…
SkillBFreshness: BClaude Code +9 - 0.521score
understand-and-refactor
nhouseholder
Join an unfamiliar product repo: map architecture and data flow, then apply the highest-ROI structural / duplication / perf / maintainability refactors.
SkillBFreshness: BClaude Code +9 - 0.521score
get-ready
nhouseholder
Session start, read-only.
SkillBFreshness: BClaude Code +9 - 0.521score
kg
nhouseholder
Keep going — audit open work, pick the highest-leverage task, implement it end-to-end, verify, then immediately take the next highest-leverage step.
SkillBFreshness: BClaude Code +9 - 0.521score
fybr
nhouseholder
Follow your best recommendation — pick one course, execute it, do not interview.
SkillBFreshness: BClaude Code +9 - 0.521score
hq-law
nhouseholder
Always-on standing law for every Cloud Agent on any ProjectsHQ repo: honesty about what was run, fail-closed on unknowns, real-money safety for betting picks…
SkillBFreshness: BClaude Code +9 - 0.521score
performance-optimize
nhouseholder
Performance engineer loop — speed, memory, scalability, extra rendering.
SkillBFreshness: BClaude Code +9 - 0.521score
wrap-up
nhouseholder
Session end.
SkillBFreshness: BClaude Code +9 - 0.517score
tailscale-cloud
nhouseholder
ALWAYS ACTIVE on Cloud Agents.
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Aug 28, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from nhouseholder
- 0.524score
whats-next
nhouseholder
Strategic next-action prioritizer: anchors on the project's north star, scores impact vs effort vs confidence vs decay, filters already-killed work, and…
SkillBFreshness: BClaude Code +9 - 0.524score
estimate
nhouseholder
Effort, time, and complexity estimation before starting a task.
SkillBFreshness: BClaude Code +9 - 0.524score
debate
nhouseholder
Runs a formal 3-round debate between Claude (Proposer) and Gemini 3.1 Pro (Challenger), ending with an AI verdict via backend API.
SkillBFreshness: BClaude Code +9 - 0.524score
reflect
nhouseholder
Session retrospective — review what was accomplished in this session versus the original goal.
SkillBFreshness: BClaude Code +9 - 0.521score
kg
nhouseholder
Keep going — audit open work, pick the highest-leverage task, implement it end-to-end, verify, then immediately take the next highest-leverage step.
SkillBFreshness: BClaude Code +9 - 0.521score
get-ready
nhouseholder
Session start, read-only.
SkillBFreshness: BClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get clean-architecture-rebuild running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
clean-architecture-rebuild skill. Agent Mart. https://agentmart.dev/skills/nhouseholder-clean-architecture-rebuild (last updated 2026-10-24)