Back to Docs
Getting Started
How a user gets value from SharpSana
This is the shortest practical path from account creation to a working product context layer across research, planning, and IDE use.
Dashboard Tabs
Projects, Settings, API Keys, Usage, Subscription, MCP for IDE.
Project Tabs
Chat, Memory, Tasks, Sprints, Roadmap, Insights, Materials, Interviews, Notifications, Settings, MCP for IDE.
01.Create account and open Dashboard
Start in sharpsana.app and open the main workspace.
- Sign up or sign in.
- Go to Dashboard and verify the left navigation.
- Main tabs: Projects, Settings, API Keys, Usage, Subscription, MCP for IDE.
02.Create your first project
Use Dashboard -> Projects -> New Project.
- Complete the 4-step wizard.
- Add project name and optional website and logo.
- Add context, files, links, markets, competitors, and optional GitHub connection.
03.Review Project Memory
Check semantic company memory and AI context.
- Open the project workspace.
- Go to Memory to review what Sharpsana currently knows about your company.
- Use chat to correct or add memories when context changes.
04.Plan execution
Turn context into concrete work.
- Tasks: create and move work items across statuses.
- Sprints: start a sprint and track completion.
- Roadmap: place initiatives on timeline and connect tasks.
05.Add evidence
Feed the system with real user and product data.
- Materials: upload docs and add links until indexed.
- Interviews: run New analysis in 3 steps.
- Use interview output to improve insights and planning.
06.Convert insights to tasks
Use Insights to prioritize and execute.
- Review new insights and approve strong candidates.
- Create one or all suggested tasks from each insight.
- Move accepted work into sprint and roadmap.
07.Optional MCP IDE setup
Connect engineering workflows to the same context.
- Generate key in Dashboard -> API Keys.
- Run npx sharpsana-cli@latest configure.
- Run npx sharpsana-cli@latest init in repository root.
Day 1 Completion Check
First sprint started
One insight converted to task
At least two materials indexed
One interview analysis created
Optional IDE MCP configured