@wix/create-app@0.0.257
Create Wix apps
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is a guarded, explicit-user-command agent configuration capability. It warrants a warning under the agent integration policy, but the inspected source does not establish malware or an unconsented attack.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
A user runs the create-app CLI and selects an MCP editor integration.
Impact
The selected editor may later execute the configured MCP client; this is an explicit project-level agent capability, not confirmed malicious behavior.
Mechanism
The generator writes an editor-specific project configuration that invokes the Wix remote MCP client through npx.
AI rationale
This is a guarded, explicit-user-command agent configuration capability. It warrants a warning under the agent integration policy, but the inspected source does not establish malware or an unconsented attack.
Evidence
build/index.jsbuild/generator-G3PPR2ZE.jstemplates/app/.mcp.json.ejs.mcp.json.cursor/mcp.json.vscode/mcp.json
Network endpoints1
mcp.wix.com/sse
Decision evidence
public snapshotAI called this Suspicious at 88.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- The package exposes a command-line entrypoint and has no install lifecycle hook in its metadata.
- The interactive creation flow asks the user to choose an MCP-enabled editor.
- The generator renders templates and writes them into the newly created project.
- The MCP template configures npx to run the Wix remote MCP client against the Wix MCP gateway.
Evidence against
- The editor choice is presented interactively, including a no-selection option, rather than being activated during installation.
- The configuration is written only as part of generating the selected project template.
Behavioral surface
ChildProcessCryptoEnvironmentVarsEvalFilesystemNetwork
HighEntropyStringsUrlStrings
NoLicense
Source & flagged code
9 flaggedbuild/index.jsView file
105__commonJS,
L106: __dirname,
L107: __esm,
...
L751: {
L752: name: "AppVeyor",
L753: constant: "APPVEYOR",
...
L1074: var vendors = require_vendors();
L1075: var env4 = process.env;
L1076: Object.defineProperty(exports, "_vendors", {
...
L1310: }
L1311: process26.exitCode = code || /* istanbul ignore next */
L1312: 0;
Critical
Download Execute
Source downloads or fetches remote code and executes it.
build/index.jsView on unpkg · L105105Trigger-reachable chain: manifest.bin -> bin/index.cjs -> bin/import.cjs -> build/index.js
L105: __commonJS,
L106: __dirname,
L107: __esm,
...
L751: {
L752: name: "AppVeyor",
L753: constant: "APPVEYOR",
...
L1074: var vendors = require_vendors();
L1075: var env4 = process.env;
L1076: Object.defineProperty(exports, "_vendors", {
...
L1310: }
L1311: process26.exitCode = code || /* istanbul ignore next */
L1312: 0;
Critical
Trigger Reachable Dangerous Capability
A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
build/index.jsView on unpkg · L10519105"node:buffer": [">= 14.18 && < 15", ">= 16"],
L19106: child_process: true,
L19107: "node:child_process": [">= 14.18 && < 15", ">= 16"],
Findings
2 Critical6 High4 Medium5 Low
CriticalDownload Executebuild/index.js
CriticalTrigger Reachable Dangerous Capabilitybuild/index.js
HighChild Processbuild/index.js
HighEvalbuild/generator-G3PPR2ZE.js
HighCredential Redirect Persistencebuild/index.js
HighCommand Output Exfiltrationbuild/index.js
HighCross File Remote Execution Contextbuild/index.js
HighSemantic Analysis Limitedbuild/generator-G3PPR2ZE.js
MediumNetwork
MediumEnvironment Vars
MediumShips Wasm Modulebuild/yoga.wasm
MediumStructural Risk Force Deep Review
LowScripts Present
Affected versions and remediation
This report applies to @wix/create-app@0.0.257.
See all versions of @wix/create-app for other recorded verdicts.
- Review the evidence and your use of @wix/create-app@0.0.257 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.