visual-ai-editor@2.1.2: Suspicious npm security report (Warn) | LPM Firewall

Loading npm security reports…

visual-ai-editor@2.1.2: Suspicious npm security report (Warn) | LPM Firewall

Source & flagged code

6 flagged · loading source

package.jsonView file

•scripts.postinstall = node dist-node/scripts/postinstall.js

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node dist-node/scripts/postinstall.js

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg

demo/.envView file

1patternName = blocked_file severity = critical matchedText = demo/.env redactedSecretContext = secretLikeLines = 1 L1: AI_API_KEY=<redacted:56 token-like>

Critical

Critical Secret

Package contains a critical-looking secret pattern.

demo/.envView on unpkg · L1

dist-node/bin/cli.jsView file

14Object.defineProperty(exports, "__esModule", { value: true }); L15: var fs = require('fs'); L16: var path = require('path');

Medium

Dynamic Require

Package source references dynamic require/import behavior.

dist-node/bin/cli.jsView on unpkg · L14

dist-node/server/index.jsView file

42Manifest entrypoint (manifest.exports) carries capability families absent from dist/build output: environment+network L42: var configLib = require('../lib/config.js'); L43: var DEFAULT_ENDPOINT = 'https://api.groq.com/openai/v1/chat/completions'; L44: var DEFAULT_MODEL = 'llama-3.3-70b-versatile'; ... L78: if (provided !== token) { L79: return res.status(401).json({ error: t('auth.invalid') }); L80: } ... L84: function assertProductionSafe(options, t) { L85: if (process.env.NODE_ENV !== 'production') L86: return; ... L105: var preset = (ai.provider && PROVIDER_PRESETS[ai.provider]) || {}; L106: var root = options.staticDir || process.cwd(); L107: var fileConfig = configLib.resolveConfig(root, options);

High

Entrypoint Build Divergence

Manifest entrypoint contains risky behavior absent from dist/build output.

dist-node/server/index.jsView on unpkg · L42

dist-node/scripts/postinstall.jsView file

2Install-time code directly mutates a foreign AI-agent control surface: L11: * `npm install` here, during development, doesn't warn about itself). L12: * - Delivers/updates AGENTS.md in the consumer project root (idempotent — L13: * see lib/agents-md.js for the merge logic). ... L28: if (result.action === 'created') { L29: console.log('[visual-ai-editor] AGENTS.md criado em ' + result.path + ' (guia para agentes de IA).'); L30: } L31: else if (result.action === 'appended' || result.action === 'updated') { L32: console.log('[visual-ai-editor] AGENTS.md atualizado em ' + result.path + '.'); L33: } ... L35: catch (e) { L36: console.warn('[visual-ai-editor] Não foi possível instalar AGENTS.md (não crítico): ' + (e && e.message)); L37: } Write operation from dist-node/lib/agents-md.js:

Critical

Ai Agent Control Hijack

Source creates an unconsented AI-agent control surface through install-time mutation or a default unauthenticated remote skill channel.

dist-node/scripts/postinstall.jsView on unpkg · L2

Source & flagged code

6 flagged

package.jsonView file

Lines 41-81json

41 "dist-node/",

42 "templates/",

43 "docs/",

44 "!docs/safe-guard/",

45 "!docs/gifs/",

46 "types/",

47 "examples/",

48 "demo/",

49 "README.md",

50 "CHANGELOG.md"

51 ],

52 "bin": {

53 "visual-ai-editor": "dist-node/bin/cli.js"

54 },

55 "scripts": {

56 "build:node": "tsc",

57 "build:browser": "tsx scripts/build.ts",

58 "build": "tsc && tsx scripts/build.ts",

59 "dev": "tsx scripts/build.ts --watch",

60 "prepublishOnly": "npm run build && npm test",

61 "postinstall": "node dist-node/scripts/postinstall.js",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L61

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg · L61

62 "test": "node --test test/*.test.js"

63 },

64 "keywords": [

65 "ai",

66 "editor",

67 "html",

68 "visual",

69 "groq",

70 "openai",

71 "opencode",

72 "anthropic",

73 "ollama",

74 "design-system",

75 "layout",

76 "component"

77 ],

78 "author": "bruno-gs-dev",

79 "license": "MIT",

80 "sideEffects": true,

81 "dependencies": {

demo/.envView file

Lines 1-2text

1AI_API_KEY=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

Critical

Critical Secret

Package contains a critical-looking secret pattern.

demo/.envView on unpkg · L1

2

dist-node/bin/cli.jsView file

Lines 1-34javascript

1#!/usr/bin/env node

2"use strict";

3/**

4 * visual-ai-editor CLI

5 *

6 * npx visual-ai-editor [start] — zero-config: interactive config on first run, then serve

7 * the project with the editor auto-injected

8 * (flags: --port <n>, --no-inject, --no-open)

9 * npx visual-ai-editor design:check — locate DESIGN.md and report section coverage

10 * npx visual-ai-editor design:init — write DESIGN.prompt.md (guided prompt for an AI agent)

dist-node/server/index.jsView file

Lines 22-62javascript

22 * any OpenAI-compatible chat-completions API works

23 * (Groq, OpenAI, OpenRouter, Ollama, LM Studio, ...)

24 *

25 * Environment fallbacks for the provider:

dist-node/scripts/postinstall.jsView file

Lines 1-22javascript

1"use strict";

2/**

3 * visual-ai-editor postinstall

4 *

5 * Best-effort, never fails `npm install`: everything here is wrapped so any

6 * error just gets swallowed (with a console.warn) instead of aborting the

7 * consumer's install.

8 *

9 * What it does:

10 * - Skips entirely when running inside this package's own repo (so