SG/Sentinel TUI Mockups: The Operator Surfaces As Acceptance Criteria

Summary

Nine ASCII-art mockups of the operator surfaces — deployment reality, live traffic, blocks, logs + S3 browsing, rules management, rule detail and testing, the deployed code, threat intel, and the deploy flow — with one principle running through all of them: every UI must be chatbot-friendly, backed by a structured TUI API, so you can talk to the surface as well as click it. The deployed-code view carries the trust punchline: "this is the exact code running in prod right now", readable because the tiny core is small.

Key concepts

• The operator surfaces — five of the mocked screens shipped in the MVP: rules, logs, blocks, status, traffic
• Chatbot-friendly by construction — the TUI API underneath is what the read-only LLM chat runs on
• The deployed-code view — sg sentinel tui status shows the exact materialised L1 engine

Key ideas

• The TUI is the visual surface; the chat is the conversational surface; both are views of the same TUI API.
• Mockups as acceptance criteria: concrete enough to implement, explicitly not final designs.
• The deferred-feature mockups (threat intel, fractal graph) were deliberately not built — they need deferred features.

On this site

The testing manual's §6–7 show these mockups made real; the try-it page documents them.

Read the document