Use your usual chat app to shape rough material into markdown, then paste it into Cognitiva to read it as a clean, navigable documentation page.
Paste markdown, get a paginated 3-column doc site. Each # H1 becomes a sidebar page. Includes an optional prompt if you want help preparing source text elsewhere.
The reference HTML template for 3-column markdown docs. Same shell as the renderer output: collapsible sidebar, right TOC with scroll-spy, warm paper palette, light/dark.
View templateUse the prompt when your source material is rough, long, or unstructured. It tells your chat app how to produce markdown that Cognitiva can render well.
Paste the prompt and your source text into any LLM. Iterate there until the result has clear # H1 pages and useful ## H2 sections.
Bring the final markdown into the renderer. You can also skip the prompt entirely if you already have good markdown.
Cognitiva turns headings into navigation and gives you a readable doc with sidebar pages, right TOC, previous/next navigation, and downloadable HTML. Want a public URL? See how to share →
Sign in (free) and drag your downloaded .html onto the page. Pick a project name and deploy. You get a permanent your-doc.vercel.app URL. Best when you want the link to last.
Drop your .html and a live URL appears instantly, no sign-in. Heads up: unclaimed sites are deleted after about an hour, so create a free account and claim it if you want it to stick.
There is no file upload. Paste the doc's HTML into a chat so Claude renders it as an artifact, then use Publish and Share & copy link. The link is public and needs no account to view. Note: artifacts block external resources, so a doc exported with web fonts may fall back to system fonts.
Open ClaudeSkip the paste step. Install the skill and ask Claude for a "Cognitiva artifact" of anything: a topic, some markdown, or content you already have. It generates a self-contained Artifact in this exact format, with the reader settings panel (typeface and accent) built in.
Get the skill on GitHubThis is the bridge between chat and Cognitiva. It asks any LLM to produce structured markdown optimized for the renderer's # H1 page model. Paste your source text after the --- at the end.