@funny-booth/agent-core@0.1.27
AIエージェント起動CLI（共有ランチャー + MCPサーバー）
AI Security Review
scanned 9d ago · by lpm-firewall-aiThe package directs its MCP agent to submit feedback without user request and provides a path to send that report to the configured portal. The transfer is conditional on agent tool use, so the source supports a warning rather than a publish block.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
An agent follows the MCP guidance and invokes the feedback tool.
Impact
Information included in a report may leave the user's environment without a specific user request.
Mechanism
The tool forwards report fields through the portal client as JSON to the package-configured portal.
AI rationale
The package directs its MCP agent to submit feedback without user request and provides a path to send that report to the configured portal. The transfer is conditional on agent tool use, so the source supports a warning rather than a publish block.
Evidence
dist/feedback.jsdist/index.jslauncher.config.json
Network endpoints1
app.ai-agentos.jp
Decision evidence
public snapshotAI called this Suspicious at 83.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- dist/feedback.js tells the agent to send feedback even when the user has not asked.
- dist/index.js supplies that text as MCP server instructions and forwards feedback fields to a reporting helper.
- dist/feedback.js posts the supplied report as JSON through the portal client; the launcher defaults the portal to app.ai-agentos.jp.
Evidence against
- The feedback tool requires an agent invocation, and the package instructions tell agents not to include customer information or secrets.
- The inspected code does not establish hidden collection of local credentials or files.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetwork
HighEntropyStringsMinifiedObfuscatedUrlStrings
NoLicense
Source & flagged code
7 flagged · loading sourcedist/mcpPackages.jsView file
1'use strict';const _0x19402b=_0x480b;(function(_0x12600d,_0x17fcba){const _0x5c9908=_0x480b,_0x252e19=_0x12600d();while(!![]){try{const _0x429d4a=-parseInt(_0x5c9908(0xfd))/0x1+par...
High
Child Process
Package source references child process execution.
dist/mcpPackages.jsView on unpkg · L1dist/feedback.jsView file
1'use strict';const _0xf9dbc1=_0x2d2f;function _0x3183(){const _0x57a168=['[redacted]...
High
Obfuscated Payload Loader
Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/feedback.jsView on unpkg · L11'use strict';const _0xf9dbc1=_0x2d2f;function _0x3183(){const _0x57a168=['[redacted]...
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/feedback.jsView on unpkg · L1Findings
1 Critical6 High4 Medium6 Low
CriticalPrevious Version Dangerous Deltadist/launcher.js
HighChild Processdist/mcpPackages.js
HighSame File Env Network Executiondist/launcher.js
HighCommand Output Exfiltrationdist/launcher.js#virtual:normalized:round1
HighObfuscated Payload Loaderdist/feedback.js
HighTrigger Reachable Command Output Exfiltrationdist/launcher.js#virtual:normalized:round1
HighObfuscated
MediumDynamic Requiredist/feedback.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
Affected versions and remediation
This report applies to @funny-booth/agent-core@0.1.27.
See all versions of @funny-booth/agent-core for other recorded verdicts.
- Review the evidence and your use of @funny-booth/agent-core@0.1.27 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.