registry / @whalent/agent-core / 0.3.495
@whalent/agent-core@0.3.495
⚠ Under reviewCore runtime for Whalent Agent
AI Security Review
scanned 19d ago · by lpm-firewall-aiDirect inspection established agent configuration capabilities but did not resolve their consent and activation paths through the heavily obfuscated CLI. Clipped source access and the entrypoint's citation verification size limit prevent a defensible clean or malicious determination.
Static reason
One or more suspicious static signals were detected.
Trigger
Tool requests can invoke skill installation; configuration helper activation remains unresolved.
Impact
These capabilities can change agent instructions and available tools; unconsented use was not established.
Mechanism
Downloaded skill files are written into platform-specific directories, and configuration helpers can replace MCP entries.
AI rationale
Direct inspection established agent configuration capabilities but did not resolve their consent and activation paths through the heavily obfuscated CLI. Clipped source access and the entrypoint's citation verification size limit prevent a defensible clean or malicious determination.
Evidence
dist/mcp.cjsdist/index.cjsconfig.toml.cursor
Decision evidence
public snapshotAI called this Manual Review at 94.0% confidence as Unknown with high false-positive risk.
AI review observations
- package.json points both the main entrypoint and CLI to dist/index.cjs; the inspected manifest contains no lifecycle hooks or runtime self-dependency.
- The flagged MCP location selects agent platform directories, including .cursor.
- dist/mcp.cjs contains a helper that modifies Codex MCP configuration in config.toml.
- dist/index.cjs dispatches tool requests to downloaded skill installation, but obfuscated activation and consent remain unresolved; source access was clipped and citation verification rejected its size.
Evidence against
- No concrete credential theft, malicious payload execution, or automatic npm install mutation was established.
- The inspected dynamic Function construction belongs to schema validation code.
Behavioral surface
CryptoDynamicRequireEnvironmentVarsEvalFilesystemNetwork
HighEntropyStringsMinifiedObfuscated
Source & flagged code
7 flagged · loading sourcedist/mcp.cjsView file
1#!/usr/bin/env node
L2: 'use strict';var a0_0x48cb70=a0_0x2b33;(function(_0x28037f,_0x203a96){var _0x38b537=a0_0x2b33,_0x35f3ef=_0x28037f();while(!![]){try{var _0x12ac1c=parseInt(_0x38b537(0x265))/0x1+-pa...
L3: || (${_0x60fd76} == "string" && ${_0x179718} && ${_0x179718} == +${_0x179718})`)[_0x5a736d(0x1801)](_0x5bb405,(0x0,_0x4209fd['_'])`+${_0x179718}`);return;case _0x5a736d(0x73f):_0x2...
High
Obfuscated Payload Loader
Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/mcp.cjsView on unpkg · L1•stage = ast_semantic_analysis; reason = ast_path_work_budget_exceeded; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/mcp.cjsView on unpkg4|| (${_0x60fd76} === "string" && ${_0x179718} && ${_0x179718} == +${_0x179718} && !(${_0x179718} % 1))`)['assign'](_0x5bb405,(0x0,_0x4209fd['_'])`+${_0x179718}`);return;case _0x5a7...
L5: || ${_0x60fd76} === "boolean" || ${_0x179718} === null`)['assign'](_0x5bb405,(0x0,_0x4209fd['_'])`[${_0x179718}]`);}}}function _0x269304({gen:_0x3134ce,parentData:_0xb2e07d,parentD...
L6: missingProperty: ${_0xcdac03},
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@0.3.495.
See all versions of @whalent/agent-core for other recorded verdicts.
- Review the evidence and your use of @whalent/agent-core@0.3.495 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.