MCP Tool
nova.sandbox_create
For browser environment, window state, and isolated execution surfaces, an agent uses this public tool to carry out the bounded step "Sandbox create". It separates target confirmation, intended state change, and verification after execution.
Canonical reference for a public MCP tool name.
- Type
- MCP tool
- Family
- Sandbox Management
- Effect
- changes state
- Status
- Explained
- Path
- 33.4
Purpose
What this entry explains
What it does
nova.sandbox_create
An agent uses this tool to carry out the bounded step "Sandbox create". It belongs to browser environment, window state, and isolated execution surfaces; agents should combine it with current discovery, target awareness, and the visible result.
Use when
- Use this entry when an agent needs to carry out the bounded step "Sandbox create" for browser environment, window state, and isolated execution surfaces.
- Use it to understand the public tool name, its expected boundary, and the response signals to check.
- Use it before chaining follow-up tools so the next step is based on current evidence.
Reference Use
How agents should cite and apply this area
Examples are maintained at family level and use only public tool names or reference paths already present in the catalog.
Family example
An agent needs to change visible state in browser environment, window state, and isolated execution surfaces without guessing target or result from an earlier session.
The agent starts with nova.sandbox_context, reads the current response or reference, and only then chooses the concrete next tool.
Change only the confirmed target, then check visible state and stop on mismatch.Contract
Inputs and important response fields
This page is a public reference. Agents and integrators should still read current MCP tool discovery before execution, because schemas can be gated by settings or version.
Inputs
name
Name for the target object. Check uniqueness before creating or changing it.
purpose
Public input field from the catalog source. Confirm the exact schema through current MCP discovery before execution.
Response fields
No fixed public response field is derived from the catalog source for this path. Use current tool discovery and the actual response before chaining the next step.
Safety
Boundary before execution
Can change browser, page, or workflow state. The target and expected result must be clear before execution.
Confirm the current target first, perform only the intended bounded action, and verify the resulting page or workflow state afterwards.
For humans, this entry names the browser or workflow state in browser environment, window state, and isolated execution surfaces that may change, so the action can be reviewed before and after execution.
High-Impact Review
Execution boundary and recheck hints
Review category: Auto-apply/mutating flows
Mutating steps may run only with confirmed target, expected effect, user control, and result check.
False assumption: a guard or auto-apply signal replaces current target and result checks.
The user must be able to review action, target, approval, block signal, and visible result.
Execute only the intended step, then verify state and stop on warning, block, or target change.
Recheck if candidate budget, gate, visible state, or expected effect do not line up.
Safety Axes
How this path can affect work
Axes are stable catalog signals for humans, agents, and LLM discovery. One path can carry several axes.
browser_state_change
Changes tab, navigation, focus, claim, scroll position, window state, or browser environment.
Confirm the target context visibly before execution and verify that the expected browser state was reached afterwards.