Generate and edit Draw.io, Mermaid, and Excalidraw diagrams from natural language using a structured JSON spec.
Coding
scaffold-mvp
Try itUse when starting a new .NET/Blazor/Angular MVP project in an empty repository.
What it does
This skill is activated **only** in brand-new repositories, right after has established the shared language, domain, and SPEC SDD.
The skill document
Scaffold MVP
Trigger
This skill is activated only in brand-new repositories, right after /grill-me-with-spec has established the shared language, domain, and SPEC SDD.
Golden Rule (Non-Negotiable)
It is strictly forbidden to build base UI components or infrastructure from scratch.
Reuse shared components and mature libraries focused on prototyping speed. This is the highest priority and non-negotiable. Always prefer shadcn/ui, Radix, MUI, or equivalent proven UI kits over custom components.
Stability and Technical Cadence (No Rushing)
MVP speed must not produce unstable code or structural shortcuts. Syntactic and architectural integrity is sovereign. The agent must follow:
- Incremental compilation check: After installing any dependency or creating a base directory, run the local build or type check command (e.g.,
npx tsc --noEmit,go build,cargo check). Never accumulate changes without confirming the current build passes. - Zero pseudo-code: Escape comments such as
// ...or// rest of the code hereare forbidden in routes or scaffold files. Every created file must be self-contained and commercially functional. - Safe dependency setup: Pin exact library versions. Always run the explicit install command to ensure clean lockfile updates (
package-lock.json,go.sum,yarn.lock,pnpm-lock.yaml). - Bridge and contract building: If the MVP depends on external services (database, auth), provide locally usable stubs or mocks. Avoid unhandled crashes on the first startup.
Workflow
Phase 1 — Context Ingestion and Stack Proposal (Aligned Bootstrap)
Do not perform generic business interrogation; the predecessor skill already established the domain.
- Read
CONTEXT.md,docs/adr/, and the approved.specs/SPEC-*.md. - Based on the discovered domain, design a hyper-productive infrastructure. Be consultative and opinionated in favor of speed.
- If the context suggests a standard web app, categorically propose the proven ecosystem: Next.js + Tailwind + shadcn/ui. For other profiles (CLI, worker, pure backend), propose the equivalent MVP stack in the respective language.
- Mandatory validation: Present the chosen stack and ask the user in Portuguese:
Baseado no nosso contexto de dominio, proponho iniciar com [STACK_ESCOLHIDA] para maxima produtividade sem reinventar a roda.
Voce concorda com esta stack ou temos alguma restricao tecnica ainda nao mapeada?
➡️ Meu palpite: concordo com a proposta.
Wait for explicit user approval. If the user wants changes, adapt the bootstrap. If the user agrees, proceed.
Phase 2 — Technical and Structural Execution
Proceed only after explicit user approval.
- Project initialization (
package.json,go.mod,pyproject.toml,Cargo.toml, etc.): configure the ecosystem. - Base stack installation (e.g.,
npx shadcn-ui@latest initwhere applicable). - Agile directory structure focused on code reuse:
/components/shared— reusable UI components injected via libraries/lib— utility functions and service integrations/hooks— custom state logic
- Generate a lean README documenting:
- local run commands
- adopted architectural view
- how rapid prototyping should be guided (reuse first)
- Run the build/type check after each significant step. Fix any error before moving on.
Stack Decision Tree
Use this table to propose a reasonable default. Confirm with the user before committing.
| Profile | Default MVP stack | UI kit |
|---|---|---|
| Web app (SaaS, marketing, dashboard) | Next.js 15 + Tailwind + React Server Components | shadcn/ui |
| CLI / script | Node.js + Commander or Python + Click / Typer | — |
| API / backend | Node.js/Express or .NET 8 or FastAPI | — |
| Mobile | React Native (Expo) or Flutter | nativewind / shadcn RN |
| Data / ML | Python + Pydantic + FastAPI or Jupyter | — |
If the user has a different preference, ask in Portuguese:
A stack padrao para este perfil e [STACK_SUGERIDA]. Voce confirma ou prefere uma alternativa?
➡️ Meu palpite: confirmo a stack padrao.
External Service Stubs
For every external dependency (database, cache, queue, auth, object storage, payment), create a local, runnable stub:
- Use Docker Compose or a local in-memory implementation where possible.
- Provide a
.env.examplewith all required keys and fake local values. - Add a health-check script.
- Never leave the app crashing on startup when a service is missing.
Return Criteria
Before handing back to the orchestrator:
-
CONTEXT.mdis updated with the chosen stack under Technical Details. -
README.mdexists with run commands and a short architecture note. - The build / type check passes cleanly.
- Lockfiles are updated and committed.
- No pseudo-code,
TODO, or// ...escape comments remain in scaffold files. - External dependencies have local stubs or mocks.
Then return control to the orchestrator reporting that the ground is ready for feature development.
Common Mistakes
| Mistake | Fix |
|---|---|
| Custom UI base components | Use a proven UI kit. |
| Skipping build checks | Run build/type check after every structural step. |
Leaving // ... comments | Every file must be self-contained and functional. |
| No local stubs for external services | Add Docker / in-memory / env-example stubs. |
| Proceeding without user stack approval | Ask and wait for explicit Portuguese confirmation. |
References
grill-me-with-spec— for producing the SPEC SDD that precedes this skillcreate-agent-harness— for installing the agent harness in the new repocreate-issues— for turning Epics into GitHub Issues
Related skills
Query Twitter/X profiles, tweets, follower events, and KOL data through the 6551 REST API.
Find why your productivity system keeps failing, then apply the smallest fix — capacity math, bottleneck routing, durable local notes.
Write, debug, and tune Playwright specs with locator strategy, trace diagnosis, and CI-aware timeouts.
Stores durable facts in a categorized, plain-markdown vault on disk, alongside your agent's built-in memory.
Join a video meeting as an AI bot with voice, avatar, and screenshare across four operating modes.
More from afonsoft
Browse all skillsSingle owner of everything under docs/architecture/ — ADRs, architecture and design documents, and architecture diagrams. Routes each deliverable to the right engine: /mermaid-architecture for Markdown-native diagrams, /drawio-architecture for editable .drawio diagrams, and the optional third-party archify skill for interactive standalone HTML diagrams (installed on demand via `npx skills add tt-a1i/archify`, only with explicit user approval). Use whenever architecture documentation, ADRs, or architecture diagrams must be created or updated.
Use when building a new MCP server in TypeScript, Python, or C# that exposes tools to LLMs.
Central entry point of the afonsoft agent harness. Use when starting a new project, resuming an existing one, planning features/Epics/releases, or running any multi-step agent-driven work. Validates and reconciles SPECs (SDD), audits the codebase and harness for gaps (security, architecture, performance, hygiene), proposes improvements, fragments work into GitHub Issues, delegates implementation/QA/review to specialized skills, and re-validates everything until delivery. Also use to review unapproved SPECs, reconcile open GitHub Issues with code, or run a final gap check before closing a release.
Use when the user asks to connect an AI agent to external apps via Composio, or when Composio CLI or MCP setup fails.
Use when initializing or migrating an AI agent harness in a repository.
Use when turning approved plans, specs, PRDs, or Epics into trackable GitHub Issues.