registry / @mode-db/server / 0.1.13

@mode-db/server@0.1.13

ModelDB WebSocket server: CLI, config loading, wire transport, and query dispatch. Depends on @mode-db/core for runtime, authoring, and wire serialization APIs.

AI Security Review

scanned 6d ago · by lpm-firewall-ai

The reachable postinstall hook writes a bundled agent instruction file into the consumer workspace. This is first-party package setup with a real agent configuration effect, but the inspected source does not establish a concrete attack.

Static reason

High-risk behavior combination matched malicious policy.

Trigger

Package postinstall.

Impact

The added file becomes guidance available to agents working in that workspace; no credential transfer or other harmful effect was established.

Mechanism

The installer creates the workspace modeldb directory and copies the bundled AGENTS.md only if no file already exists.

AI rationale

Evidence

package.jsonscripts/install-modeldb-workspace.mjstemplates/modeldb/AGENTS.md

Decision evidence

public snapshot

AI called this Suspicious at 89.0% confidence as Dangerous Capability with medium false-positive risk.

Evidence for warning

The package runs a postinstall script that adds bundled guidance to the consumer workspace.
The script writes to modeldb/AGENTS.md only when that file is absent.
The bundled file provides ModeDB development guidance and advises treating connectToken as secret.
The source inspected does not establish credential collection, network transfer, or destructive behavior.

Evidence against

The installer exits without replacing an existing modeldb/AGENTS.md.
The separate postuninstall hook removes the product cache directory.

Behavioral surface

Source

EnvironmentVarsFilesystemNetwork

Supply chain

HighEntropyStringsMinified

ManifestNo manifest risk signals triggered.

scanned 29 file(s), 85.0 KB of source

Source & flagged code

5 flagged

package.jsonView file

Lines 47-70json

47 "commander": "^12.1.0",

48 "fastify": "^5.8.4",

49 "ws": "^8.18.0",

50 "@mode-db/common": "^0.1.10",

51 "@mode-db/core": "^1.0.13",

52 "@mode-db/shared": "^0.1.12"

53 },

54 "publishConfig": {

55 "access": "public"

56 },

57 "devDependencies": {

58 "@types/node": "^22.10.0",

59 "@types/ws": "^8.18.1",

60 "typescript": "^5.9.0"

61 },

62 "scripts": {

63 "prebuild": "pnpm -C ../.. run build:shared && pnpm -C ../.. run build:common && pnpm -C ../.. run build:core",

64 "pretypecheck": "pnpm -C ../.. run build:shared && pnpm -C ../.. run build:common && pnpm -C ../.. run build:core",

65 "build": "tsc -p tsconfig.json",

66 "typecheck": "tsc --noEmit -p tsconfig.json",

67 "postinstall": "node ./scripts/install-modeldb-workspace.mjs",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L67

68 "postuninstall": "node ./scripts/clear-modeldb-cache.mjs"

69 }

70}

Medium

Suspicious Lifecycle Evidence

The package runs a postinstall script that adds bundled guidance to the consumer workspace.

package.jsonView on unpkg · L67

scripts/install-modeldb-workspace.mjsView file

Lines 1-31javascript

1/**

2 * Copies AI/workspace guidance into **`./modeldb/AGENTS.md`** at the consuming project root.

3 * Uses **`INIT_CWD`** (npm/pnpm) when present so installs from nested **`node_modules`** still target the app workspace.

4 */

5import { copyFileSync, existsSync, mkdirSync } from 'node:fs';

6import { dirname, join } from 'node:path';

Source & flagged code

5 flagged · loading source

package.jsonView file

•scripts.postinstall = node ./scripts/install-modeldb-workspace.mjs

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

67"postinstall": "node ./scripts/install-modeldb-workspace.mjs", "postuninstall": "node ./scripts/clear-modeldb-cache.mjs" } }

Medium

Suspicious Lifecycle Evidence

The package runs a postinstall script that adds bundled guidance to the consumer workspace.

package.jsonView on unpkg · L67

scripts/install-modeldb-workspace.mjsView file

1/** L2: * Copies AI/workspace guidance into **`./modeldb/AGENTS.md`** at the consuming project root. L3: * Uses **`INIT_CWD`** (npm/pnpm) when present so installs from nested **`node_modules`** still target the app workspace. L4: */ L5: import { copyFileSync, existsSync, mkdirSync } from 'node:fs'; L6: import { dirname, join } from 'node:path'; ... L10: const pkgRoot = join(__dirname, '..'); L11: const template = join(pkgRoot, 'templates', 'modeldb', 'AGENTS.md'); L12: const rawInit = process.env.INIT_CWD?.trim(); ... L14: const destDir = join(cwd, 'modeldb'); L15: const destFile = join(destDir, 'AGENTS.md'); L16:

Findings

1 Critical1 High6 Medium4 Low

CriticalAi Agent Control Hijackscripts/install-modeldb-workspace.mjs

HighInstall Time Lifecycle Scriptspackage.json

MediumNetwork

MediumEnvironment Vars

MediumStructural Risk Force Deep Review

MediumSuspicious Lifecycle Evidencepackage.json

MediumAi Review Evidencescripts/install-modeldb-workspace.mjs

MediumSuspicious Lifecycle Evidencescripts/install-modeldb-workspace.mjs

LowNon Install Lifecycle Scripts

LowScripts Present

LowFilesystem

LowHigh Entropy Strings

Affected versions and remediation

This report applies to @mode-db/server@0.1.13.

See all versions of @mode-db/server for other recorded verdicts.

Review the evidence and your use of @mode-db/server@0.1.13 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: 2026-10-17 04:45:34.396Z (UTC).

Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.