registry / @whalent/agent-core / 0.3.516
@whalent/agent-core@0.3.516
Core runtime for Whalent Agent
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe inspected code supports skill installation into agent configuration directories through a named tool, which leaves a concrete agent extension capability. It does not establish an automatic install-time attack, credential theft, or unrestricted file write.
Static reason
One or more suspicious static signals were detected.
Trigger
The skill-management tool receives an install action and a skill slug.
Impact
Installed skill content can affect later agent behavior; path checks limit writes to the selected skill directory.
Mechanism
The package downloads skill files using the account token, then writes them into the selected platform and scope directory.
AI rationale
The inspected code supports skill installation into agent configuration directories through a named tool, which leaves a concrete agent extension capability. It does not establish an automatic install-time attack, credential theft, or unrestricted file write.
Evidence
dist/mcp.cjs.claude/skills.cursor/skills.agents/skills
Network endpoints1
memory.whale
Decision evidence
public snapshotAI called this Suspicious at 83.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The named skill-management tool accepts an install action and slug, downloads the skill using the account token, and passes the returned files to the installer.
- The installer selects a scope-specific skill directory under the supported agent platform directories and writes downloaded files there.
- The writer resolves each file path beneath the selected skill directory and skips paths that escape it.
- The skill download helper uses the package’s configured service base and sends the account token to its skill endpoint.
- The manifest identifies the runtime entrypoints and contains no npm lifecycle scripts.
Evidence against
- The agent configuration change occurs through an explicit skill-management tool action; the inspected source does not show it running from an npm install hook.
- File-path checks constrain downloaded writes to the selected skill directory, and the package manifest has no lifecycle hook.
Behavioral surface
CryptoDynamicRequireEnvironmentVarsEval
Source & flagged code
7 flagged · loading sourcedist/mcp.cjsView file
1#!/usr/bin/env node
L2: 'use strict';var a0_0x1633af=a0_0x51a5;(function(_0x4d3b09,_0x27c549){var _0x659c83=a0_0x51a5,_0x450e02=_0x4d3b09();while(!![]){try{var _0x34358d=-parseInt(_0x659c83(0xe85))/0x1*(-...
L3: || (${_0x51b2c6} == "string" && ${_0x4f3da8} && ${_0x4f3da8} == +${_0x4f3da8})`)[_0x3621b0(0x10b4)](_0x3a923a,(0x0,_0xb47c43['_'])`+${_0x4f3da8}`);return;case _0x3621b0(0x222c):_0x...
High
Obfuscated Payload Loader
Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/mcp.cjsView on unpkg · L1•stage = ast_semantic_analysis; reason = ast_path_work_budget_exceeded; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/mcp.cjsView on unpkg4|| (${_0x51b2c6} === "string" && ${_0x4f3da8} && ${_0x4f3da8} == +${_0x4f3da8} && !(${_0x4f3da8} % 1))`)[_0x3621b0(0x10b4)](_0x3a923a,(0x0,_0xb47c43['_'])`+${_0x4f3da8}`);return;ca...
L5: || ${_0x51b2c6} === "boolean" || ${_0x4f3da8} === null`)['assign'](_0x3a923a,(0x0,_0xb47c43['_'])`[${_0x4f3da8}]`);}}}function _0x37a5cc({gen:_0x3c180d,parentData:_0x46dd48,parentD...
L6: missingProperty: ${_0x23d3eb},
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@0.3.516.
See all versions of @whalent/agent-core for other recorded verdicts.
- Review the evidence and your use of @whalent/agent-core@0.3.516 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.