registry / @whalent/agent-core / 0.3.500
@whalent/agent-core@0.3.500
Core runtime for Whalent Agent
AI Security Review
scanned 13d ago · by lpm-firewall-aiThe package manifest has no install lifecycle hook, but the bundled runtime includes network-backed skill installation into agent configuration directories. The inspected source does not resolve consent or destination controls, supporting a warning for agent extension lifecycle risk rather than a block.
Static reason
One or more suspicious static signals were detected.
Trigger
A caller invokes the bundled skill installation flow.
Impact
Depending on caller controls and destination configuration, installed skill files could alter the agent’s available instructions or capabilities.
Mechanism
The runtime downloads skill file data through an authenticated request and writes the supplied files under a platform-specific agent skill directory.
AI rationale
The package manifest has no install lifecycle hook, but the bundled runtime includes network-backed skill installation into agent configuration directories. The inspected source does not resolve consent or destination controls, supporting a warning for agent extension lifecycle risk rather than a block.
Evidence
dist/mcp.cjs
Decision evidence
public snapshotAI called this Suspicious at 79.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The package identifies itself as Whalent Agent’s core runtime and exposes a bundled MCP entrypoint.
- The bundled code maps agent platforms to their configuration directories, including `.claude` and `.agents`.
- The bundle contains a skill installer that creates a destination directory and writes supplied files there.
- The bundle downloads skill files through an authenticated request to a configured base URL.
Evidence against
- The manifest declares no npm lifecycle scripts; the agent configuration writes are in the bundled runtime.
- The inspected source does not establish whether skill installation requires user consent or who controls the download base URL.
Behavioral surface
CryptoDynamicRequireEnvironmentVarsEvalFilesystemNetwork
HighEntropyStringsMinifiedObfuscatedUrlStrings
Source & flagged code
7 flagged · loading sourcedist/mcp.cjsView file
1#!/usr/bin/env node
L2: 'use strict';var a0_0x2891b3=a0_0x1b80;(function(_0x2b9020,_0x3e66fc){var _0x32ce92=a0_0x1b80,_0x2f48a6=_0x2b9020();while(!![]){try{var _0x13108f=-parseInt(_0x32ce92(0x111c))/0x1*(...
L3: || (${_0x4f75ba} == "string" && ${_0x146107} && ${_0x146107} == +${_0x146107})`)[_0x5529b8(0x359)](_0x5cca0a,(0x0,_0x21efd7['_'])`+${_0x146107}`);return;case'integer':_0x4ea071[_0x...
High
Obfuscated Payload Loader
Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/mcp.cjsView on unpkg · L1•stage = ast_semantic_analysis; reason = ast_path_work_budget_exceeded; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/mcp.cjsView on unpkg4|| (${_0x4f75ba} === "string" && ${_0x146107} && ${_0x146107} == +${_0x146107} && !(${_0x146107} % 1))`)['assign'](_0x5cca0a,(0x0,_0x21efd7['_'])`+${_0x146107}`);return;case'boolea...
L5: || ${_0x4f75ba} === "boolean" || ${_0x146107} === null`)['assign'](_0x5cca0a,(0x0,_0x21efd7['_'])`[${_0x146107}]`);}}}function _0x2a666c({gen:_0x569d6f,parentData:_0x2c43e7,parentD...
L6: missingProperty: ${_0x4fcb9c},
Medium
Findings
3 High7 Medium5 Low
HighObfuscated Payload Loaderdist/mcp.cjs
HighObfuscated
HighSemantic Analysis Limiteddist/mcp.cjs
MediumDynamic Requiredist/mcp.cjs
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperdist/jupyter_sidecar.py
MediumOversized Source Filedist/index.cjs
MediumOversized Cli Entrypointdist/index.cjs
MediumStructural Risk Force Deep Review
LowEvaldist/mcp.cjs
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowNo License
Affected versions and remediation
This report applies to @whalent/agent-core@0.3.500.
See all versions of @whalent/agent-core for other recorded verdicts.
- Review the evidence and your use of @whalent/agent-core@0.3.500 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.