@whalent/agent@0.3.375
Stable wrapper for Whalent Agent core runtime
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is suspicious because high-impact remote-management behavior is hidden behind obfuscation, so its safety cannot be verified from the published source. It is not malicious under the install-hook policy because no automatic registry-install trigger or concrete attack behavior was found.
Static reason
One or more suspicious static signals were detected.
Trigger
An explicit user invocation of the whalent CLI, particularly daemon mode.
Impact
A user cannot independently audit the exact remote-command and update behavior before running it.
Mechanism
Obfuscated remote-managed agent daemon with subprocess execution and helper updating.
AI rationale
This is suspicious because high-impact remote-management behavior is hidden behind obfuscation, so its safety cannot be verified from the published source. It is not malicious under the install-hook policy because no automatic registry-install trigger or concrete attack behavior was found.
Evidence
dist/index.cjsREADME.md
Decision evidence
public snapshotAI called this Suspicious at 91.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The executable entrypoint is heavily obfuscated, preventing normal audit of its runtime behavior.
- The entrypoint loads child-process support and launches processes with inherited environment variables.
- Package text describes a daemon that receives remote commands and downloads and activates helper manifests.
- The package is a CLI agent that can start managed AI workers after an explicit command.
Evidence against
- package.json has no preinstall, install, or postinstall hook.
- The only lifecycle script is prepublishOnly, which is not run for registry consumers.
- No concrete credential theft, destructive action, or unconsented install-time mutation was established from the inspected source.
Behavioral surface
DynamicRequireEnvironmentVarsFilesystem
HighEntropyStringsMinifiedPossibleObfuscationTrivial
Source & flagged code
5 flagged · loading sourcedist/index.cjsView file
1#!/usr/bin/env node
L2: 'use strict';(function(_0x96b9e0,_0x6ff8e){const _0x2e172d=_0x96b9e0();function _0x3604db(_0x1653c5,_0x194646,_0x2e337e,_0x15f095){return a0_0x28fe(_0x1653c5-0x15e,_0x2e337e);}func...
High
Obfuscated Payload Loader
Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/index.cjsView on unpkg · L11#!/usr/bin/env node
L2: 'use strict';(function(_0x96b9e0,_0x6ff8e){const _0x2e172d=_0x96b9e0();function _0x3604db(_0x1653c5,_0x194646,_0x2e337e,_0x15f095){return a0_0x28fe(_0x1653c5-0x15e,_0x2e337e);}func...
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/index.cjsView on unpkg · L1dist/jupyter_sidecar.pyView file
•path = dist/jupyter_sidecar.py
kind = build_helper
sizeBytes = 18590
magicHex = [redacted]
Medium
Ships Build Helper
Findings
1 High6 Medium5 Low
HighObfuscated Payload Loaderdist/index.cjs
MediumDynamic Requiredist/index.cjs
MediumEnvironment Vars
MediumShips Build Helperdist/jupyter_sidecar.py
MediumOversized Source Filedist/core.cjs
MediumOversized Cli Entrypointdist/core.cjs
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowNo License
Affected versions and remediation
This report applies to @whalent/agent@0.3.375.
See all versions of @whalent/agent for other recorded verdicts.
- Review the evidence and your use of @whalent/agent@0.3.375 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.