Ngenaire Docs

What the AI Assistant can do

The Ngenaire AI Assistant is a multi-turn chat agent with deep knowledge of your project. It can read the artifacts you have already authored and create or modify artifacts on your behalf — all under your direction. It is available everywhere in the app.

Where to open it

You can reach the assistant two ways:

  1. AI Console panel. A side panel that slides in over your current page. Open it with the AI button at the right of the top bar, or press Cmd/Ctrl+K. Press Esc to close it. The panel stays available while you navigate, so it is ideal for "ask about what I'm looking at."
  2. AI Assistant page. A dedicated full-page chat at /ai-assistant, best for longer working sessions.

Both surfaces share the same conversation context for the active project.

What it can do

The assistant works through the platform's tools, so it can act across the whole systems-engineering lifecycle. It can:

  • Read any artifact you can read and answer questions about your project.
  • Draft and decompose requirements, and rewrite them in EARS form.
  • Generate diagrams (UML / SysML, Data Flow Diagrams, PlantUML).
  • Identify, score, and draft mitigations for risks.
  • Recommend verification methods and draft test cases and test procedures.
  • Draft SOW and CONOPS sections.
  • Propose schedule tasks and surface critical-path and slip risk.
  • Generate status briefs, design-review summaries, and reports.

How it works

Live streaming

The assistant streams its work as it goes: you see its thinking, the actions it is taking, and the results, in real time. You can follow along rather than waiting for a single block of text at the end.

Deep links to artifacts

When the assistant creates or changes something, it ends its reply with deep links — one per artifact — that take you straight to that artifact's page in the app. Clicking a link navigates in-app (no full reload, the chat stays open) and opens the artifact so you can review exactly what was created or changed.

Reads Project Memory

The assistant reads Project Memory at the start of a turn, so durable facts — decisions, constraints, stakeholder preferences — carry forward across conversations without you repeating them.

Confirms before bulk or destructive changes

When a request implies multiple creates or any delete, the assistant first describes its plan in plain English and asks you to confirm before it proceeds.

Permissions

The assistant respects your role on the current project — it cannot do anything you could not do yourself, and it never reaches outside your project or organization:

  • Viewer — can read artifacts and answer questions, but cannot create or modify anything.
  • Editor and Owner — full read/write capability on the active project.

See Roles and permissions.

Credits

AI actions consume credits from your account balance. See your billing and credits for how credits are granted and tracked.

Going further

Related