Human-Agent Control Surfaces and Operator Experience
Design interfaces that let people understand, direct, interrupt, and judge autonomous work without reading raw logs.
A focused view of boundaries, contracts, state, authority, failure paths, and tradeoffs drawn from this chapter.
Reconstruct and defend this chapter’s architecture.
Reconstruct the architecture, name each boundary, and defend the tradeoffs.
Open the source exercise
Design one screen for a high-risk WorkOrder awaiting release. Include intent, plan deviation, current artifact, failed evidence, pending approval, rollback, safe actions, and raw trace drill-down. Explain the information hierarchy.
4. Tradeoffs and alternatives
More detail increases transparency and cognitive load. Use progressive disclosure: decision summary first, trace and artifacts on demand. Chat is flexible and poor at showing parallel state and evidence lineage. Structured interfaces are clear and can feel rigid; allow conversational clarification without bypassing records.
5. Current Mission Control Implementation
The current material specifies operator screens for Missions, plans, attempts, evidence, approval, review, release, health, and learning. It also distinguishes human workflow preferences from authority.
The curriculum does not yet provide one complete interaction model covering plan preview, live progress, pause/resume, intervention, notification, review inbox, accessibility, and user-research measures. Current pages should be evaluated against this system rather than treated as sufficient because they expose records.
Review this chapter.
Challenge a claim, boundary, missing failure mode, unclear term, or unsupported evidence statement.
- Claim
- Boundary
- Failure
- Evidence