Scroll is a real-time collaborative wiki for product, engineering, QA, design, sales, and ops, anyone who works out of a shared knowledge base. Everyone gets AI writing, semantic search, diagrams, and decks. Teams on SimplifyQA also get live test cases, execution reports, and traceability embedded as first-class blocks.
Every other doc tool added AI as a sidebar. Scroll was designed around an MCP-powered AI server from day one, the same backbone that powers writing also powers semantic search, diagram generation, presentation creation, and Q&A grounded in your real content.
The only doc platform with native test case, execution report, traceability matrix, defect analysis, and dashboard blocks. Not iframes, not screenshots, real React components with live SimplifyQA data.
Single sign-on, shared users, shared projects, shared roles. A SimplifyQA admin does not onboard Scroll separately, their team is already there. Switching between testing and documentation is one click.
Not embeds. Not screenshots. Not iframes. Live React components rendered inside the editor, pulling real data from SimplifyQA in real time.
Preconditions, steps, expected results, data variables, version, and live pass/fail status, rendered as a first-class React component.
Five-stage drill-down from execution to items to runs to steps to failed-step detail. Live. Never a screenshot.
Auto-generated from linked requirements, test cases, and defects. Coverage analysis. Orphan detection. Zero spreadsheets.
Structured RCA: category, root cause, contributing factors, corrective actions, with live status pulled from SimplifyQA.
Live environment status, configuration, dependencies, and version info embedded right next to the runbooks that reference them.
10+ live SimplifyQA metric widgets, execution status, pass rate trend, defect count, release health, traceability coverage.
Multi-LLM. Per-org daily cost caps. Per-user rate limiting. Secret detection on every input. Prompt-injection wrapping. Full audit log. Enterprise AI, not a sidebar chatbot.
Generate, improve, rewrite, summarize (short / bullets / TLDR / executive), translate with formatting preserved.
Rewrite in formal, casual, friendly, professional, technical, or neutral tone, and adjust length independently.
pgvector-powered embeddings that find documents by meaning, not just keywords.
Ask questions of a single page, an entire space, or a connected repository, with citations to the source content.
Describe a system in plain English, get a Mermaid flowchart, sequence diagram, ER diagram, or mindmap.
Turn any page or space into a slide deck with 15+ web-native layouts. Export to PDF or PPTX.
Connect a GitHub / GitLab / Bitbucket repo and generate architecture docs, API references, and onboarding guides from the code.
OpenAI, Anthropic, Google Gemini, Azure OpenAI. Per-org cost caps. Per-user rate limits. Full audit trail.
Every page [[wiki-links]] to every other. Scroll draws the map, a live graph of what clusters, what bridges, and what's floating alone. Semantic search and grounded Q&A ride the same graph your team writes.
Type [[ and pull in any page. Every link becomes a live edge, renamed on both sides when the target moves.
Every page shows who cites it. No manual index. Hubs surface themselves; orphans stop hiding.
Drag, pin, zoom, filter by tag. Clusters show what your team actually writes about together.
Ask a KB and get an answer with citations, sourced from the same graph you can see.
Every doc tool got the wiki part right. Scroll gets the QA part right too, and pairs it with an MCP-native AI backbone the others cannot retrofit.
Same SSO, same users, same organizations, same projects. A user already in SimplifyQA is already in Scroll. Auto-refreshing tokens. Permission cache invalidated on any role change. Auth is invisible to the user.
Every page linked, every hub surfaced, every test, defect, and release live inside the same knowledge graph your team writes.