Set up in 5 minutes

New workspaces open a short guided setup in the dashboard (Set up BLAZLE, also in the sidebar). It takes you from sign-up to your first blocked request. You can leave at any step and pick up later.

1. Where do your AI agents work? Pick agents and scripts (API), GitHub, MCP clients, or VS Code and Cursor. Only what you pick is shown next.
2. Connect. Create an API key with a ready-to-copy example, connect GitHub in one click, or create an MCP token with the client configuration filled in.
3. Your company policies. Drop your security policies (PDF, Word, Markdown, HTML or text) and turn on ready-made rules. Rules warn first. See Policy Library.
4. Protection level. Balanced (recommended) blocks critical risks and prompt injections and redacts secrets in prompts. Strict blocks anything rated High or above. The AI prompt-injection check is included on every plan.
5. See it work. Run a harmless test command. BLAZLE blocks it and the record appears in Intercepts.

Owners, Admins and Security Analysts can complete setup. Everyone else sees the progress.