claude-multiacc@2.0.3

Unified Claude Code and OpenAI Codex subscription pooling with quota-aware selection.

AI Security Review

scanned 15d ago · by lpm-firewall-ai

The global postinstall launches installer behavior that can add command shims and scheduled maintenance, creating a meaningful agent CLI lifecycle risk. The reviewed code supports the package’s account-pooling purpose and provides no concrete evidence of malicious activity.

Static reason

One or more suspicious static signals were detected.

Trigger

Global npm installation, unless CI or the skip variable is set.

Impact

The package can intercept those CLI invocations and run recurring maintenance; inspected evidence did not establish credential exfiltration or other malicious behavior.

Mechanism

Postinstall invokes the package installer, which can create command shims and scheduled account checks.

AI rationale

Evidence

scripts/postinstall.mjsinstall.shbin/claude/usr/local/bin/claude

Decision evidence

public snapshot

AI called this Suspicious at 87.0% confidence as Dangerous Capability with medium false-positive risk.

Evidence for warning

The manifest runs a postinstall script.
On global installs, postinstall invokes the package installer; that installer can replace command lookup with package shims and add scheduled account checks.
The `claude` shim is designed to shadow and dispatch the real CLI, a package-aligned but meaningful command interception capability.

Evidence against

The postinstall skips local and CI installs and supports an explicit skip variable.
The selector test file contains ordinary unit-test cases, and the selector entrypoint processes JSON input without shell execution.

Behavioral surface

Source

ChildProcessEnvironmentVarsFilesystemShell

Supply chain

HighEntropyStrings

ManifestNo manifest risk signals triggered.

scanned 2 file(s), 6.67 KB of source

Source & flagged code

7 flagged

package.jsonView file

Lines 34-62json

34 "gowalk"

35 ],

36 "author": {

37 "name": "Gowalk"

38 },

39 "license": "MIT",

40 "repository": {

41 "type": "git",

42 "url": "git+https://github.com/gowalk-public/claude-multiacc.git"

43 },

44 "homepage": "https://github.com/gowalk-public/claude-multiacc#readme",

45 "bugs": {

46 "url": "https://github.com/gowalk-public/claude-multiacc/issues"

47 },

48 "publishConfig": {

49 "access": "public"

50 },

51 "engines": {

52 "node": ">=18"

53 },

54 "scripts": {

55 "postinstall": "node scripts/postinstall.mjs",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L55

Medium

Ambiguous Install Lifecycle Script

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

package.jsonView on unpkg · L55

56 "test": "bash tests/run-tests.sh && python3 tests/test_selector.py"

Medium

Suspicious Lifecycle Evidence

The manifest runs a postinstall script.

package.jsonView on unpkg · L54

57 },

58 "dependencies": {

59 "update-notifier": "^7.3.1"

60 }

61}

install.shView file

•path = install.sh kind = build_helper sizeBytes = 21240 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

install.shView on unpkg

tests/test_selector.pyView file

•path = tests/test_selector.py kind = payload_in_excluded_dir sizeBytes = 11717 magicHex = [redacted]

High

Payload In Excluded Dir

Package hides binary, compressed, or executable-looking payloads in test/fixture/hidden paths.

tests/test_selector.pyView on unpkg

bin/multiacc-selectView file

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

bin/multiacc-selectView on unpkg

scripts/postinstall.mjsView file

Lines 1-33javascript

1#!/usr/bin/env node

2// Runs after `npm install`. On a GLOBAL install/update it wires the addon up by

3// running install.sh (idempotent), so `npm i -g claude-multiacc@latest` fully updates

4// the shim, CLI, and scheduled jobs with no extra step. Deliberately best-effort:

5// npm's own CI, sandboxes, and non-global/dev installs must never be broken by it, and

6// a failed install.sh must not fail `npm install` — it only prints how to finish by hand.

7import { spawnSync } from 'node:child_process';

8import { fileURLToPath }

Source & flagged code

7 flagged · loading source

package.jsonView file

•scripts.postinstall = node scripts/postinstall.mjs

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node scripts/postinstall.mjs

Medium

Ambiguous Install Lifecycle Script

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

package.jsonView on unpkg

54"scripts": { "postinstall": "node scripts/postinstall.mjs", "test": "bash tests/run-tests.

Medium

Suspicious Lifecycle Evidence

The manifest runs a postinstall script.

package.jsonView on unpkg · L54

install.shView file

Findings

3 High6 Medium3 Low

HighInstall Time Lifecycle Scriptspackage.json

HighPayload In Excluded Dirtests/test_selector.py

HighSemantic Analysis Limitedbin/multiacc-select

MediumAmbiguous Install Lifecycle Scriptpackage.json

MediumEnvironment Vars

MediumShips Build Helperinstall.sh

MediumStructural Risk Force Deep Review

MediumSuspicious Lifecycle Evidencepackage.json

MediumSuspicious Lifecycle Evidencescripts/postinstall.mjs

LowScripts Present

LowFilesystem

LowHigh Entropy Strings

Affected versions and remediation

This report applies to claude-multiacc@2.0.3.

See all versions of claude-multiacc for other recorded verdicts.

Review the evidence and your use of claude-multiacc@2.0.3 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: 2026-10-08 16:16:42.786Z (UTC).

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

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Runtime surface

Behavioral surface

Source & flagged code