spirewise 1.0.3 → 1.6.0

package/skills/nvidia-inception-starter/SKILL.md

@@ -0,0 +1,179 @@
+---
+name: nvidia-inception-starter
+description: >-
+  Expert startup advisor that prepares, evaluates, and optimizes a startup for
+  acceptance into NVIDIA Inception. Use when the user asks to "assess my startup
+  for NVIDIA Inception", "am I ready for NVIDIA Inception", "improve my NVIDIA
+  application", "NVIDIA Inception readiness", or wants a scored audit + roadmap.
+  Researches NVIDIA Inception and accepted startups, scores the company 0–100,
+  audits website/product/AI/architecture/NVIDIA-alignment, and writes a full
+  readiness report to an `nvidia-inception/` folder in the project root.
+---
+
+# NVIDIA Inception Starter
+
+Prepare a startup for NVIDIA Inception: research the program and its accepted
+companies, score the startup, audit every dimension that affects acceptance, and
+produce a brutally honest readiness report with a 90-day action plan.
+
+## Persona
+
+Act as a **former NVIDIA Inception evaluator, AI startup investor, accelerator
+mentor, and deep-tech strategist**. Be evidence-based, data-driven, highly
+critical, and brutally accurate, but always actionable. Hold both an investor
+mindset and a technical mindset. Never inflate scores to be nice — a realistic,
+specific critique is more valuable than encouragement.
+
+## Inputs needed
+
+Gather (ask, or infer from the repo/website/links the user provides):
+- Company name, website, one-line description, incorporation status + age
+- What it builds, target market, industry, business model, pricing
+- AI/ML usage and category (CV, GenAI, robotics, physical AI, enterprise AI…)
+- Technical stack and any NVIDIA technologies used
+- Funding stage, traction (users, revenue, growth, partnerships), team/founders
+- Pitch deck / homepage copy if available
+
+If details are missing, mark them as `[UNKNOWN]` in the report and lower the
+relevant scores — do not invent facts.
+
+## Step 1 — Research (do this before scoring)
+
+Research to current, real sources. Do not rely on memory alone.
+- **Official program:** eligibility, application process, tiers, benefits, FAQ
+  (nvidia.com/startups and the Inception FAQ).
+- **Accepted startups:** study a broad sample (aim for many, ideally 50–100+)
+  from NVIDIA startup showcases, member spotlights, case studies, and blogs.
+- For each accepted startup, note: industry, problem, market, product/AI
+  category, funding stage, tech stack, NVIDIA tech used, homepage headline,
+  value prop, traction, and competitive edge.
+- **Detect patterns:** most common categories, AI use cases, funding stages,
+  founder backgrounds, website structures, messaging, and NVIDIA-tech usage.
+- Convert patterns into **acceptance signals** (what accepted startups share)
+  and **risk signals** (what rejected/ineligible profiles share).
+
+## Verified program facts (ground truth — keep current by re-checking)
+
+Eligibility (typical):
+- **Free, equity-free, no fees.** No application deadline; rolling review
+  (usually a few business days to ~2 weeks).
+- Must be **incorporated** (registered legal entity), have a **working company
+  website**, use a **business email** (no generic/alias addresses), and employ
+  **at least one developer**.
+- Generally **under 10 years old**, and **building AI / ML / data-science /
+  accelerated-computing products** — not selling consulting or outsourced dev.
+- **Open to all funding stages** (bootstrapped → growth).
+
+Commonly **not eligible**: consulting / outsourcing dev shops, crypto
+(mining/trading), pure cloud service providers, resellers/distributors, and
+public companies.
+
+Tiers (startups progress as they grow): **Emerging → Scale → Premier**. Higher
+tiers unlock more cloud credits, deeper engineering support, DLI training,
+co-marketing, and investor access.
+
+Benefits: NVIDIA SDKs + developer tools, Deep Learning Institute (DLI) training,
+preferred hardware pricing, partner **cloud credits**, **Inception Capital
+Connect / VC Alliance** investor access, co-marketing, badges, and events.
+
+NVIDIA technologies to align with (cite the ones that genuinely fit): CUDA,
+cuDNN, TensorRT, Triton, RAPIDS, NIM, NeMo, Riva, Metropolis, Holoscan,
+Omniverse, Isaac (robotics), Jetson (edge), DGX / NVIDIA AI Enterprise.
+
+## Step 2 — Score the startup (0–100)
+
+Score each category, multiply by weight, sum to a total. Always justify each
+score with specific evidence and a one-line "to reach the next level" note.
+
+| Category | Weight |
+|---|---|
+| AI Strength | 20 |
+| Technical Innovation | 15 |
+| NVIDIA Alignment | 15 |
+| Market Opportunity | 10 |
+| Team Quality | 10 |
+| Product Maturity | 10 |
+| Website Quality | 5 |
+| Pitch Readiness | 5 |
+| Scalability | 5 |
+| Competitive Advantage | 5 |
+
+Map the total to an **acceptance probability** band and state it plainly:
+- 85–100 — Very strong; likely accept. Polish and apply now.
+- 70–84 — Strong; address 2–3 gaps, then apply.
+- 50–69 — Borderline; meaningful work needed before applying.
+- < 50 — Not ready; follow the roadmap first.
+
+NVIDIA Alignment scoring guide: highest when the startup genuinely trains/runs
+GPU-accelerated AI and would use NVIDIA tech in production; low for thin
+"AI wrapper" products with no real compute need. Reward authenticity, not
+buzzwords.
+
+## Step 3 — Audits
+
+Run each and list concrete findings + fixes:
+- **Website audit:** homepage headline, value proposition, product + AI
+  explanation, technical credibility, use cases, case studies, architecture
+  clarity, visual quality, trust signals, investor readiness. Compare against
+  top Inception / fast-growing AI startups.
+- **Product audit:** maturity (idea → MVP → production), differentiation, moat.
+- **AI readiness audit:** real AI vs. wrapper, data advantage, model strategy,
+  evaluation/quality, compute needs.
+- **Technical architecture audit:** scalability, GPU utilization, inference
+  path, edge vs. cloud, MLOps.
+- **NVIDIA alignment audit:** which NVIDIA technologies fit and how to adopt
+  them credibly in the roadmap.
+
+## Physical / Embodied AI module (if relevant)
+
+If the startup touches robotics, autonomous/industrial systems, vision AI, edge
+AI, digital twins, simulation, agentic, or embodied AI, recommend the fitting
+stack: **Isaac** (robotics), **Omniverse** (simulation/digital twins),
+**Jetson** (edge), **CUDA/TensorRT** (perf), **NIM** (inference microservices),
+**NVIDIA AI Enterprise**. Tie each recommendation to a concrete use case — never
+list tech for its own sake.
+
+## Output — write the report file
+
+1. Create an `nvidia-inception/` folder in the project root if missing.
+2. Write `nvidia-inception/inception-readiness-report.md` (UTF-8 Markdown) with:
+
+```
+# NVIDIA Inception Readiness — <Company>
+Generated: <YYYY-MM-DD>
+
+## Executive Summary
+## NVIDIA Inception Acceptance Score   (XX/100, with per-category table)
+## Acceptance Probability               (band + one-line verdict)
+## Strength Analysis
+## Weakness Analysis
+## Competitive Analysis
+## Website Audit
+## Product Audit
+## AI Readiness Audit
+## Technical Architecture Audit
+## NVIDIA Alignment Audit
+## Improvement Roadmap                   (prioritized, effort vs. impact)
+## 90-Day Action Plan                    (Days 0–30 / 31–60 / 61–90, owners)
+## Application Readiness Report           (eligibility checklist + go/no-go)
+```
+
+Include the eligibility checklist explicitly (incorporated, website, business
+email, ≥1 developer, <10 yrs, AI product not consulting, not in an excluded
+category) and flag any blocker that would cause an automatic rejection.
+
+## Quality bar
+
+- Evidence over opinion: cite the accepted-startup patterns your scores rely on.
+- Be specific and quantified; replace vague advice with concrete next actions.
+- Brutally honest: if it's not ready, say so and explain exactly why.
+- Recommend NVIDIA tech only where it authentically fits the product.
+
+## Verification checklist (run before finishing)
+
+1. `nvidia-inception/inception-readiness-report.md` exists with every section.
+2. Category scores sum correctly to the 0–100 total; each has evidence.
+3. Acceptance probability band matches the total score.
+4. Eligibility checklist is complete; any auto-reject blocker is flagged.
+5. Every recommendation is specific, prioritized, and actionable.
+6. No invented facts — unknowns are marked `[UNKNOWN]`, not fabricated.

package/skills/nvidia-product-inventor/README.md

@@ -0,0 +1,66 @@
+# nvidia-product-inventor
+
+Reads your **existing startup-idea files** and invents **3–6 concrete products**
+(not marketing copy) engineered to fit NVIDIA Inception. Each product gets a
+short brandable name (**4–10 letters**) and an explicit NVIDIA stack —
+**physical/compute hardware (GPUs, Jetson, DGX…) first**, then SDKs, then
+software.
+
+> Builds on **`nvidia-inception-idea-booster`** (preference levers) and
+> **`nvidia-inception-starter`** (acceptance lens). Products only — no website
+> or landing-page copy.
+
+## Install
+
+```bash
+npx spirewise install -s nvidia-product-inventor          # pick agents + scope
+npx spirewise install -s nvidia-product-inventor -a claude,cursor -sc workspace
+```
+
+## Remove
+
+```bash
+npx spirewise remove -s nvidia-product-inventor           # pick agents + scope
+npx spirewise remove -s nvidia-product-inventor -a claude,cursor -sc both
+```
+
+(No Node? `./install.sh -s nvidia-product-inventor` and `./install.sh remove -s nvidia-product-inventor`.)
+
+## Use
+
+Put your idea notes anywhere in the project, then ask your agent:
+
+> "Invent products for NVIDIA Inception from our idea"
+
+It reads those files and writes a **`products_raw/`** folder with one subfolder
+per product.
+
+## Output
+
+```
+products_raw/
+├── _overview.md            # all products: name, letters, one-liner, primary GPU
+├── <Product1>/
+│   ├── product.md
+│   ├── logo-brief.md
+│   └── <Product1>-logo-1080x1080.png
+├── <Product2>/…
+└── …                       # 3–6 products total
+```
+
+Each `product.md` covers the problem, what it does, **why GPU compute is
+essential**, the NVIDIA stack (hardware → SDKs → software), physical/edge angle,
+moat, market, Inception fit, and MVP scope. Each product also ships a **real
+1080×1080 PNG logo** on a solid background.
+
+## Rules it enforces
+
+- **3–6 products**, each in its own folder under `products_raw/`.
+- Product names are **4–10 letters, letters only**, brandable; folder name = product name.
+- NVIDIA stack ordered **physical/GPU first**, then SDKs, then software.
+- Every product must genuinely need accelerated compute (no AI wrappers).
+- Each product gets a **proper designed logo** — 1080×1080 PNG, **solid background**
+  (primary/white/black/other, no transparency or gradient), a wordmark/lettermark/
+  combination mark (not an SVG/UI icon or emoji), saved in the product's folder.
+
+See `SKILL.md` for the full method and the per-product template.

package/skills/nvidia-product-inventor/SKILL.md

@@ -0,0 +1,167 @@
+---
+name: nvidia-product-inventor
+description: >-
+  Read a startup's existing idea/notes files and invent 3–6 concrete PRODUCTS
+  (not marketing copy) engineered to fit NVIDIA Inception preferences. Each product
+  gets a short brandable name (4–10 letters) and an explicit NVIDIA stack —
+  physical/compute hardware (GPUs, Jetson, DGX…) first, then SDKs, then software.
+  Use when the user asks to "invent products", "ideate products for NVIDIA",
+  "what should we build for Inception", or "turn our idea into products". Writes a
+  `products_raw/` folder with one subfolder per product. Builds on
+  nvidia-inception-idea-booster (preference levers) and nvidia-inception-starter
+  (acceptance lens). Products only — no website/landing copy.
+---
+
+# NVIDIA Product Inventor
+
+Turn the startup's idea into **3–6 buildable products** that maximize NVIDIA
+Inception fit. This is a product/ideation skill — invent *what to build*, not
+marketing copy.
+
+## Uses the sibling skills
+
+- Apply the **NVIDIA-preference levers** from `nvidia-inception-idea-booster`
+  (deep tech, physical/agentic AI, GPU-essential, defensibility, big market,
+  ecosystem fit) when shaping every product.
+- Pre-screen each product with the **`nvidia-inception-starter`** acceptance lens
+  (AI strength, technical innovation, NVIDIA alignment, market, moat). Keep only
+  products that would plausibly score well; discard weak/"AI-wrapper" ones.
+
+## Step 1 — Read the existing idea
+
+Discover and read the startup files in the project: `idea*`, `README*`, `pitch*`,
+`business-plan*`, `vision*`, `notes*`, `one-pager*`, `concept*`, `startup*`,
+`docs/`, etc. Restate the problem, users, market, and any tech in your own words.
+If too little exists, ask 3–5 focused questions first. Mark gaps `[UNKNOWN]` —
+never fabricate. Stay anchored to the founder's domain and intent.
+
+## Step 2 — Invent 3–6 products
+
+Generate **between 3 and 6** product concepts that each:
+- Solve a sharp, real problem within the startup's domain.
+- **Require GPU/accelerated compute** to exist (a must-have, not nice-to-have).
+- Lean into **physical / edge / agentic AI** where credible.
+- Have a defensible moat (proprietary data/model/hardware or NVIDIA integration).
+
+### Product naming rules (enforce strictly)
+- **4–10 letters**, letters only (A–Z), no spaces, digits, or symbols.
+- Brandable and coined/short (e.g., `Vexel`, `Roboq`, `Synapt`, `Orbix`).
+- Count the letters and confirm each name is within 4–10 before using it.
+- The product folder name = the product name.
+
+## NVIDIA stack for each product (order: physical → SDK → software)
+
+List concrete, fitting choices — **lead with physical/compute hardware (GPUs)**:
+
+1. **Physical / compute hardware (GPUs first):** data-center GPUs (H100, H200,
+   A100, GB200/Blackwell), workstation GPUs (RTX), **Jetson** (Orin/Nano) for
+   edge/robotics, **DGX** systems, **Grace Hopper / Grace Blackwell**, IGX/Holoscan
+   dev kits, robotics platforms on Jetson.
+2. **SDKs / libraries:** CUDA, cuDNN, **TensorRT**, **Triton** Inference Server,
+   RAPIDS, **Isaac** (Sim / ROS / Lab) for robotics, **Omniverse** for
+   simulation/digital twins, **Metropolis** for vision, **Holoscan** for sensor/edge,
+   **NeMo** for LLMs, **Riva** for speech, DeepStream, cuOpt, Modulus, Parabricks
+   (genomics), Morpheus (cybersecurity), Maxine.
+3. **Software / platforms:** **NVIDIA AI Enterprise**, **NIM** inference
+   microservices, NeMo framework/microservices, Omniverse Cloud, Fleet Command,
+   Base Command.
+
+Pick only what authentically fits each product, and say *why* each piece is used.
+
+## Output — `products_raw/` with one folder per product
+
+1. Create `products_raw/` in the project root if missing.
+2. For each product, create `products_raw/<ProductName>/` and write
+   `products_raw/<ProductName>/product.md`:
+
+```
+# <ProductName>
+Letters: <n> (must be 4–10)
+Tagline: <one line, what it is>
+Generated: <YYYY-MM-DD>
+
+## Problem
+## Target user
+## What it does (product, not marketing)
+## Why GPU / accelerated compute is essential
+
+## NVIDIA Stack
+### Physical / compute hardware (GPUs first)
+- <GPU / Jetson / DGX / …> — <why>
+### SDKs / libraries
+- <CUDA / TensorRT / Isaac / Omniverse / …> — <why>
+### Software / platforms
+- <NVIDIA AI Enterprise / NIM / …> — <why>
+
+## Physical AI / Edge angle        (if applicable)
+## Defensibility / moat
+## Market & wedge
+## Inception fit                    (1–2 lines: why this scores well)
+## Build notes / MVP scope
+## Logo                             (see logo-brief.md + <ProductName>-logo-1080x1080.png)
+## Open questions / [UNKNOWN]
+```
+
+3. Also write `products_raw/_overview.md` listing every product: name, letter
+   count, one-liner, and primary NVIDIA hardware.
+
+## Logo for each product (required)
+
+Every product gets a **real, designed brand logo** — not an SVG icon, UI/material
+icon, emoji, or clip-art glyph. Treat it like a brand identity deliverable.
+
+### Design (research first)
+- Briefly research 3–5 reference logos in the product's space; note the visual
+  conventions, then design something distinct (don't copy).
+- Choose a logo type that fits the short name: **wordmark** (the full name in
+  custom-styled type), **lettermark/monogram** (initials), or **combination mark**
+  (a simple custom symbol + the name). Coherent, ownable, memorable.
+- Define a tiny brand spec: palette (hex), typography style, and the mark concept.
+
+### Hard output rules
+- Format **PNG**, exactly **1080 × 1080 px**.
+- **Solid background** — a single flat color with NO transparency and NO gradient.
+  Use the brand **primary color**, or white, or black, or another solid color that
+  gives strong contrast with the mark. State the exact bg hex in the brief.
+- The logo must be a proper logo (typographic wordmark/lettermark/combination
+  mark), centered with safe padding — not a tiny icon on a big canvas.
+- Save it **inside that product's folder**:
+  `products_raw/<ProductName>/<ProductName>-logo-1080x1080.png`
+- Also write `products_raw/<ProductName>/logo-brief.md` with: logo type, concept,
+  background hex, foreground/accent hex, typography, and references consulted.
+
+### How to generate it
+1. **Preferred:** if an image-generation tool/model is available, prompt it for a
+   *professional brand logo* (specify: wordmark/lettermark, the exact name, color
+   palette, solid background, flat vector-style, no photographic texture), then
+   ensure the saved file is a 1080×1080 PNG on a solid background.
+2. **Fallback (always works):** render a deliberate typographic logo
+   programmatically — e.g. Python + Pillow — drawing the wordmark/monogram in a
+   chosen font and color on a solid-color 1080×1080 canvas, with optional simple
+   geometric accent. This must look designed (kerning, weight, composition), not
+   like a stock icon. Verify dimensions and that the background is fully solid.
+
+Do **not** ship: transparent PNGs, gradients, generic icon-font glyphs, stock
+emoji, or screenshots. One clean branded logo per product.
+
+## Quality bar
+
+- **Products, not copy.** No website/landing/marketing text — concrete product
+  concepts and technical plans only.
+- Every product must be GPU-essential and NVIDIA-aligned, with the stack ordered
+  physical → SDK → software.
+- Names obey the 4–10 letter rule (verified) and folder names match.
+- Stay within the founder's domain; elevate, don't fabricate.
+
+## Verification checklist (run before finishing)
+
+1. Read the existing idea files (and reflected the founder's domain).
+2. `products_raw/` exists with **3–6** product subfolders, each with `product.md`.
+3. Every product name is **4–10 letters, letters only**, and matches its folder.
+4. Each product lists NVIDIA hardware (GPU/physical) first, then SDKs, then software.
+5. Each product justifies why GPU/accelerated compute is essential.
+6. `products_raw/_overview.md` summarizes all products.
+7. Every product folder has a **1080×1080 PNG logo on a solid (non-transparent,
+   non-gradient) background**, plus `logo-brief.md` — a real wordmark/lettermark/
+   combination mark, not an SVG/UI icon or emoji. Verify file dimensions.
+8. No marketing/website copy; unknowns marked `[UNKNOWN]`, not invented.