UISandbox
NPM · UISANDBOX-MCP
Play with your app's design in the browser: load a build, turn the knobs, export the code.
Available components
Last scanned (2 hours ago)
Scores higher than 72% of npm servers
Category Browser Automation
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security98
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- No install/post-install scripts declared.Pass
- 30 of 94 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency48
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 14 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability79
- 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Pass
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 1375 tokens (~152/item across 9 items; 9 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management100
- No destabilizing schema changes in the last 30 days.Pass
Tool Coverage78
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 35% of tool parameters carry a description.Partial
Tool Safety75
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- 0 of 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation; "wait_for_send" implies "send" and declares no destructiveHint at all, which the MCP spec reads as destructive by default. See how to fix → Fail
- An AI judge read all 10 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 4 Nov 26 0
- Stability: 0.97 → pass security
- 3 Nov 26 0
- Stability: pass → 0.97 functional
- 28 Oct 26 0
- Stability: 0.97 → pass security
- 27 Oct 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 93 to 97. That category is still filling its 30-day observation window: 28 days of observed history at the previous scan, 29 at this one. The score rises as the window fills, whether or not the server changes.
- 26 Oct 26 −1
- Stability: pass → 0.93 functional
- 22 Oct 26 +11
- Known CVEs: unverified → pass ▲ security
- Dependency health: unverified → 0.85 ▲ functional
- 21 Oct 26 −11
- Stability: pass → unverified ▼ security
- Known CVEs: pass → unverified ▼ security
- Tool safety: pass → unverified ▼ security
- Malware scan: pass → unverified ▼ security
- Capabilities: pass → unverified ▼ functional
- Dependency health: 0.85 → unverified ▼ functional
- Tool coverage: 100 → unverified ▼ functional
- Schema quality: 100 → unverified ▼ functional
- Package version: 0.5.0 → 0.5.3 functional
- Package version: 0.5.0 → 0.5.2 functional
- Package version: 0.5.0 → 0.5.1 functional
- 20 Oct 26 +26
- Malware scan: unverified → pass ▲ security
- Known CVEs: unverified → pass ▲ security
- Stability: 0.97 → pass security
- Dependency health: unverified → 0.85 ▲ functional
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 5 Nov 2026 · Analysed npm/uisandbox-mcp@0.5.3
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | npm |
Background: How many MCP packages publish verified provenance →
Dependencies 94 packages
| Packages resolved | 94 |
|---|---|
| Stale | 30 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
export Export ~76
Export the current state: sheet-css (your values as CSS custom properties), sheet-json, patch (literal → value list), tokens-css / tokens-json / tailwind / shadcn (the --k-* design tokens), swift, android-xml, android-kotlin.
| Name | Type | Req | Description |
|---|---|---|---|
| format | string | yes | – |
| id | string | yes | – |
No output schema declared.
No examples provided.
load Load a built web app ~217
Load a web app: a RUNNING dev server by its local address (devUrl, e.g. http://localhost:5173 — live mode: no build needed, hot reload passes through, best while the user is developing), or a BUILT app (a zip of its dist/build/out folder, or a repo zip carrying the build) by URL or local path. Returns the project id, screens, tokenised values and the baseline the knobs stand on. Refuses a build that cannot render (iOS/Android/source-only) with the reason. After a devUrl load, `open` it: the tokenised values and the knobs arrive in the browser as the app renders.
| Name | Type | Req | Description |
|---|---|---|---|
| devUrl | string | – | The address of a running dev server on this machine, e.g. http://localhost:3000 |
| zipPath | string | – | Local path to the zip |
| zipUrl | string | – | URL of the zip (any URL the server can fetch, incl. GitHub /archive/…zip) |
No output schema declared.
No examples provided.
open Open the sandbox ~180
Open UISandbox in the user's browser with this project loaded — the real, interactive sandbox: the whole app rendered 1:1, every knob there to turn by hand. Use this whenever the user wants to SEE or PLAY with the design themselves ("can I look at this in a sandbox", "let me try some changes"), not only read numbers. Served from 127.0.0.1 by this process; nothing leaves the machine. What the user turns in the browser flows back here: a later `export`, `verify` or `screenshot` uses it.
| Name | Type | Req | Description |
|---|---|---|---|
| carry | string | – | The knobs the user turned on a kit at uisandbox.org, as the prompt they copied there names it: offered to them when the app opens |
| id | string | yes | – |
| screen | string | – | Screen path to open first |
No output schema declared.
No examples provided.
screens List screens ~25
The screens (HTML entries) of a loaded project.
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | – |
No output schema declared.
No examples provided.
screenshot Screenshot ~46
A PNG of the screen as it renders with the current knobs (the sandbox frame only).
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | – |
| screen | string | – | – |
| width | integer | – | – |
No output schema declared.
No examples provided.
set Turn the knobs ~64
Apply knobs to a loaded project. Every dial has ×1 (or 0) = as in their code; colours are #rrggbb. Returns what moved (count and a sample of value → value).
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | – |
| knobs | object | yes | – |
No output schema declared.
No examples provided.
state Current knobs ~36
The knobs as they stand now — including what the user turned by hand in an opened sandbox (and when).
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | – |
No output schema declared.
No examples provided.
verify Check 1:1 ~91
Run the 1:1 check in a real browser: the untouched build vs. the tokenised build with the identity sheet, computed styles of every element diffed. Also returns the reach meter. This is the same check a visitor runs on uisandbox.org.
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | – |
| screen | string | – | Screen path (from screens); default the first |
| width | integer | – | – |
No output schema declared.
No examples provided.
wait_for_send Wait for the user to send it back ~119
After `open`: wait until the user presses "Send to your agent" in the sandbox, then return what they turned (from → to, in the panel's own names), how many values moved, and the patch that writes it into their source. Returns at once if they already sent and you have not collected it. If the time runs out first it says the sandbox is still open: call it again (or ask them).
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | – |
| timeoutSeconds | integer | – | How long to wait (default 240) |
No output schema declared.
No examples provided.
How do I install the UISandbox MCP server?
UISandbox runs locally as an npm package, launched with npx -y uisandbox-mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · uisandbox-mcp
claude mcp add ideelab-uisandbox -- npx -y uisandbox-mcp
What is the UISandbox MCP server?
UISandbox is an MCP server listed in the public MCP registry as io.github.Ideelab/uisandbox. Play with your app's design in the browser: load a build, turn the knobs, export the code. This page covers its npm package (uisandbox-mcp).
Is the UISandbox MCP server safe to use?
UISandbox scores 82 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 5 November 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the UISandbox MCP server expose?
UISandbox exposes 9 tools: load, screens, set, export, open, and 4 more. Their descriptions and schemas cost roughly 854 tokens of context every time the server is loaded.
Is the UISandbox MCP server still maintained?
UISandbox is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the UISandbox MCP server under?
UISandbox declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.