Infographics: five pages, WebP conversion, sidebar group icons

README.md

@@ -8,7 +8,7 @@ Drop your pitch deck, market research, business plan, competitor notes and call
 
 > Every company, customer, metric and transcript used as an example in this repo is invented. Nothing here is drawn from any real client engagement.
 
-[**New to working with AI**](docs/ai-basics.md) · [**What the words mean**](docs/what-things-are.md) · [**The method**](docs/method.md) · [**Build the knowledge base**](docs/knowledge-base.md) · [**Prompt library**](prompts/INDEX.md) · [**For coaches**](docs/for-coaches.md) · [**For programmes**](docs/for-portfolios.md) · [**Quickstart**](QUICKSTART.md)
+[**New to working with AI**](docs/ai-basics.md) · [**What the words mean**](docs/what-things-are.md) · [**The infographics**](docs/infographics.md) · [**The method**](docs/method.md) · [**Build the knowledge base**](docs/knowledge-base.md) · [**Prompt library**](prompts/INDEX.md) · [**For coaches**](docs/for-coaches.md) · [**For programmes**](docs/for-portfolios.md) · [**Quickstart**](QUICKSTART.md)
 
 [![Read it as a site](https://img.shields.io/badge/read%20it%20as%20a%20site-fritzhand.github.io%2Fstartup--stack-0b5a86?style=flat-square)](https://fritzhand.github.io/startup-stack/)
 ![No code required](https://img.shields.io/badge/code%20required-none-0E6B63?style=flat-square)
@@ -63,11 +63,11 @@ There is a second reader, and the repo now serves them too: the **programme** ru
 | [`stack/`](stack/) | **The templates.** Ten numbered sections — company, customer, market, product, GTM, operations, money, capital, brand, pulse — one markdown file each, with front matter. [`stack/INDEX.md`](stack/INDEX.md) is the router an AI reads first, and [`stack/CONTEXT.md`](stack/CONTEXT.md) is the one-pager it reads next. This *becomes* your knowledge base; you fill it in, you do not copy it out. |
 | [`prompts/`](prompts/INDEX.md) | **The prompt library.** 19 prompts: bootstrap, gap scan, competitive intelligence, unit economics, the list of 100, the pitch deck, fundraise readiness, the weekly recap, meeting-to-actions, and the adversarial review that tells you what an investor will attack. Each opens with a `Requires:` line and stops if what it needs is missing. Most want a filled-in stack; [`00`](prompts/00-bootstrap-the-stack.md), [`14`](prompts/14-meeting-to-actions.md) and [`15`](prompts/15-scrape-a-site.md) run against an empty one, which is how a stack starts to fill. The last three belong to `portfolio/` and never touch a founder's stack. |
 | [`worksheets/`](worksheets/) | **The repeatable artifacts.** Fill-in files you produce over and over: startup-prep, customer interview, competitor profile, two-week experiment, SOP entry, trade/order sheet, advisor scope letter, one-pager, list of 100, weekly recap. |
-| [`docs/`](docs/) | **The method, in thirteen pages.** New to any of this: [AI basics](docs/ai-basics.md) and [what the words mean](docs/what-things-are.md) — the terminal, an editor, Git, GitHub, connectors, model tiers — each in plain language with a diagram. The method: [why it works](docs/method.md), [building the knowledge base](docs/knowledge-base.md), [the front-matter schema](docs/front-matter.md), [getting your material in](docs/getting-material-in.md) source by source, and [website to inbox](docs/scraping.md). Running it: [the weekly recap](docs/weekly-recap.md), [for coaches](docs/for-coaches.md), [the safety rules](docs/safety.md) for giving an AI access to your files, and [on a schedule](docs/automation.md) — recipes for Microsoft 365, Google Workspace, Zoom, Slack, and none of them. Two are for programmes only: [for programmes](docs/for-portfolios.md) and [the exchange folder](docs/exchange.md). |
+| [`docs/`](docs/) | **The method, in eighteen pages.** New to any of this: [AI basics](docs/ai-basics.md) and [what the words mean](docs/what-things-are.md) — the terminal, an editor, Git, GitHub, connectors, model tiers — each in plain language with a diagram. The method: [why it works](docs/method.md), [building the knowledge base](docs/knowledge-base.md), [the front-matter schema](docs/front-matter.md), [getting your material in](docs/getting-material-in.md) source by source, and [website to inbox](docs/scraping.md). Running it: [the weekly recap](docs/weekly-recap.md), [for coaches](docs/for-coaches.md), [the safety rules](docs/safety.md) for giving an AI access to your files, and [on a schedule](docs/automation.md) — recipes for Microsoft 365, Google Workspace, Zoom, Slack, and none of them. Two are for programmes only: [for programmes](docs/for-portfolios.md) and [the exchange folder](docs/exchange.md). And [the infographics](docs/infographics.md) — thirty pictures across four pages, covering [what AI is](docs/infographics-ai.md), [the tools](docs/infographics-tools.md), [the method](docs/infographics-method.md) and [running a portfolio](docs/infographics-portfolio.md); take them for a workshop. |
 | [`portfolio/`](portfolio/) | **Optional — for a programme, not a founder.** The templates an incubator, accelerator or studio uses to run this across many companies: a router, a 1-to-5 maturity rubric per function, the cross-company theme scan, and a company folder to copy. Delete it if you are a founder. See [docs/for-portfolios.md](docs/for-portfolios.md). |
 | [`_inbox/`](_inbox/) | Where raw material lands before it is processed. Git-ignored by default. |
 | [`tools/`](tools/) | **Optional.** One script that turns a website into citable markdown in `_inbox/`. You never need it — it saves an hour of copy-and-paste. See [docs/scraping.md](docs/scraping.md). |
-| [`web/`](web/) | **This project's own site**, generated from the markdown in this repo. Not part of your stack — delete it when you fork. |
+| [`web/`](web/) | **This project's own site**, generated from the markdown in this repo, plus [`web/infographics/`](web/infographics) — the thirty images the infographics pages show. Not part of your stack — delete it when you fork, but take the infographics first if you want them. |
 
 ## Quickstart
 

docs/infographics-ai.md

@@ -0,0 +1,59 @@
+# Infographics — what AI is
+
+Eight pictures covering the foundations: what a language model actually is, how one gets built, what people mean by weights, why the models keep improving, and what changes when the same model is put in a different place. Nothing here is specific to this repository. Take any of it for a workshop.
+
+Click an image to open it at full size.
+
+---
+
+## What this thing actually is
+
+The plain description, with no metaphor that misleads. A model predicts the next piece of text, repeatedly. Everything useful about it and everything dangerous about it follow from that one sentence.
+
+![A plain description of a large language model: it predicts the next piece of text, one piece at a time. What follows from that — fluency, format, patience — and what also follows: it produces the plausible continuation rather than the true one.](../web/infographics/what-this-thing-actually-is.webp)
+
+## How a model gets made
+
+Four stages — gather, pre-train, fine-tune, align — and the reason a new version arrives faster than the cost of the first one would suggest.
+
+![The four stages of building a language model, from an enormous cleaned corpus through pre-training, fine-tuning on good conversations, and alignment against human preference.](../web/infographics/how-a-model-gets-made.webp)
+
+## What weights actually mean
+
+The part of a model that *is* the model: a very large list of numbers, none of them written by a person, none of them readable afterwards. This is also where open weights and closed weights get explained, which is the choice that matters if you work under a confidentiality obligation.
+
+![Weights explained as the numbers on the connections in a network, adjusted trillions of times during training. Open weights versus closed weights, and why the trade is capability against control.](../web/infographics/what-weights-actually-means.webp)
+
+## The loops that make models improve
+
+Four feedback cycles — human preference, AI feedback, verifiable rewards, synthetic data — and an honest note on where each one stops working.
+
+![Four self-improvement loops in model training, and the limits of each: no compiler exists for 'is this good strategic advice'.](../web/infographics/the-loops-that-make-models-improve.webp)
+
+## The same model in four different places
+
+Website, desktop app with connectors, editor, API. The model is identical in all four. What differs is what it can see and what it is allowed to do — and that difference is most of the reason two people get different results from the same question.
+
+![One model shown across four surfaces — web chat, desktop app with connectors, an IDE, and the API — with what each one can see and what it may do.](../web/infographics/the-same-model-four-places.webp)
+
+## What agentic actually means
+
+The difference between asking for text and delegating a task, and the inner loop — plan, act, observe, check — that makes the second one possible.
+
+![Assistant versus agent, and the plan-act-observe-check loop at the centre of agentic work, with the three things it requires: tools, visibility of results, and a stopping condition.](../web/infographics/what-agentic-actually-means.webp)
+
+## Which model should I use
+
+Three tiers, a separate dial for thinking effort, and the rule of thumb that saves the most money: start in the middle and move for a reason.
+
+![Model tiers from small and fast to largest, the separate effort dial, and the context window explained as the desk the model works at.](../web/infographics/which-model-should-i-use.webp)
+
+## Where your words actually go
+
+What happens between pressing enter and getting an answer, and why it explains the three things that puzzle people most: rising cost in long threads, forgetting the start of a conversation, and why an index is a cost decision rather than a tidiness preference.
+
+![The path from a typed question through context assembly, tokenisation, the model, and streamed output — and why a long conversation gets slower and dearer with every turn.](../web/infographics/where-your-words-actually-go.webp)
+
+---
+
+Next: [the tools](infographics-tools.md) — terminal, Git, Node, databases and the rest of the words.

docs/infographics-method.md

@@ -0,0 +1,79 @@
+# Infographics — the method
+
+Nine pictures of what this repository actually describes. Together they are the whole method: how scattered files become a knowledge base, how it stays honest, how material keeps arriving, and what you do with it every week.
+
+Click an image to open it at full size.
+
+---
+
+## The Startup Stack method
+
+The whole thing on one page. Raw material on the left, ten sections in the middle, the work that comes out on the right, and the three loops that keep it alive.
+
+![Scattered company files becoming a ten-section knowledge base, and the work that comes out of it — with the three loops: build, enrich, pulse.](../web/infographics/the-startup-stack-method.webp)
+
+The written version is [the method](method.md).
+
+## How the AI answers without reading everything
+
+Why an index makes a knowledge base cheaper as it grows rather than dearer. This is the single idea that most people miss, and it is the reason the whole thing is structured the way it is.
+
+![A question routed through the rules file and a one-line-per-file index to two sections, against the alternative of re-reading everything on every question.](../web/infographics/how-ai-answers-without-reading-everything.webp)
+
+## The ten sections
+
+The fixed spine every stack shares, the dependencies between them, and the part that is genuinely yours to decide.
+
+![The ten sections from company to pulse, with dependency arrows, and the distinction between the fixed numbering and the files you put inside each one.](../web/infographics/the-ten-sections.webp)
+
+The written version is [build the knowledge base](knowledge-base.md).
+
+## How every fact gets tagged
+
+The quality system that stops an AI inventing your numbers. Four tags, who is allowed to apply each, the one-way promotion only a human may perform, and the fifth tag that exists because judgement is not a checkable claim.
+
+![The four fact tags with who may apply each, the human-only promotion arrow, and the assessment tag held separately.](../web/infographics/how-every-fact-gets-tagged.webp)
+
+## Front matter, line by line
+
+Eleven fields annotated with the one thing that breaks when each is wrong. Thirty seconds to write; the difference between a router and a pile.
+
+![A real front-matter block with leader lines out to what each field does and what breaks without it.](../web/infographics/front-matter-line-by-line.webp)
+
+The written version is [the front matter schema](front-matter.md).
+
+## Getting your material in
+
+Three routes from wherever your company lives today into one folder, and the list of what stays out whichever route you take.
+
+![Export by hand, connectors and website scraping, all converging on one inbox folder, with the material that stays out of all three.](../web/infographics/getting-your-material-in.webp)
+
+The written version is [getting your material in](getting-material-in.md).
+
+## Putting the pull on a schedule
+
+Connecting a subscription to mail and drive without mirroring your inbox — the setup, the weekly cycle, the daily-or-weekly question, and the monthly audit that most people skip.
+
+![Connector setup, a weekly pull cycle landing in the inbox folder, the daily versus weekly trade, and the monthly audit.](../web/infographics/putting-the-pull-on-schedule.webp)
+
+The written version is [running it on a schedule](automation.md).
+
+## Private master, shared derived
+
+How one stack produces four different documents without leaking. Three sensitivity bands, four outputs, and one band that never leaves the machine at all.
+
+![A stack in three sensitivity bands feeding four different outputs, with the restricted band drawn outside the boundary entirely.](../web/infographics/private-master-shared-derived.webp)
+
+The written version is [the safety rules](safety.md).
+
+## The weekly recap
+
+Thirty minutes, six sections, and the only part of this system that compounds. Including the arrow that brings last week's unmet commitments back as this week's uncomfortable question.
+
+![The six sections of a weekly recap as a loop, with last week's commitments returning as this week's 'what did not move', and the carve for each audience.](../web/infographics/the-weekly-recap.webp)
+
+The written version is [the weekly recap](weekly-recap.md).
+
+---
+
+Next: [running a portfolio](infographics-portfolio.md) — the layer for a programme rather than a single founder.

docs/infographics-portfolio.md

@@ -0,0 +1,35 @@
+# Infographics — running a portfolio
+
+Three pictures of the layer that sits above a single founder: what an incubator, accelerator, studio or university programme keeps for itself, and how it does that without holding anybody's master copy.
+
+Click an image to open it at full size. The written version is [for programmes](for-portfolios.md).
+
+---
+
+## Running this across a portfolio
+
+The three-zone architecture. The founder's stack on the left under their control, the programme's record on the right built from what coaches already produce, and a deliberately thin shared folder between them.
+
+Including the constraint that defeats most attempts at this: you almost certainly hold no equity, so any system that depends on founder discipline is accurate for about three weeks.
+
+![Three zones — the founder's stack, a two-door shared folder, and the programme's record — with the note that neither side's master lives in the middle.](../web/infographics/running-this-across-a-portfolio.webp)
+
+## Every session becomes two files
+
+Why the factual record and the coach's read must never be the same document, and the test that decides which one a sentence belongs in: said out loud in the room, or thought and not said.
+
+![One coaching session forking into a factual record for the founder and an attributed read for the coach team, with the divider test between them.](../web/infographics/every-session-becomes-two-files.webp)
+
+The written version is [the exchange folder](exchange.md).
+
+## What stage are they really at
+
+Five maturity levels scored on evidence rather than on funding, revenue or headcount — and the one step a programme can actually cause.
+
+![Five levels from asserted to managed, with the jump from written to evidenced highlighted, and the list of things a level is not.](../web/infographics/what-stage-are-they-really-at.webp)
+
+The written version is [the maturity rubric](../portfolio/rubric.md).
+
+---
+
+Back to [all the infographics](infographics.md).