WxO Builder
The AI review rates the findings as likely false positive, but the risk score (51/100) still counts them. The review covers version 1.6.6, not the analysed version 1.7.5.
Analysis record
- Analysed
- 6 months ago 27 April 2026
- Version
- v1.7.5
- Artifact
- SHA256 c4ff22fb…fa9b84
- Detail
- Full list
Evidence ledger
Ranked by severity · findings with a source location link to the code viewer
Publisher Evidence
Limited evidenceMarkus van Kempen
Publisher identity, store signals, distribution reach, and warning signals used for context. Treat this as supporting evidence, not a clean bill of health.
13 evidence rows available.
Finding Categories
AI Security Report
AI Security Review
Evidence context: threat category none; evidence quality weak.
The WxO Builder extension (vscode, v1.6.6) by MarkusvanKempen has no security findings in its evidence bundle. Its stated purpose—managing IBM Watsonx AI agents, tools, and flows—justifies filesystem read/write access for file operations and process execution for local testing. The absence of YARA code-smell findings (e.g., postinstall_payload_execution, credential_theft) confirms no malicious postinstall behavior or credential exfiltration. No IoC hits on suspicious domains or IPv6 fragments were detected, eliminating supply chain poisoning concerns. Credential-access findings are absent, indicating no targeting of sensitive files like .env or SSH keys. The strongest counterargument is its low user count (27), but this alone does not indicate malicious intent without corroborating evidence such as suspicious permissions or network calls. Absence of exfiltration patterns, obfuscation in non-locale files, or bundled dependency anomalies further supports benign behavior. The extension's functionality aligns with legitimate development tool requirements, and no evidence suggests excessive capability beyond its stated purpose. Additionally, the lack of findings in dist/ or bundle files rules out multiplicative false positives from bundled dependencies. The extension's network calls, if any, are likely to official servers for AI agent management, not exfiltration endpoints. Workspace access is limited to necessary files for AI tool development, with no evidence of broad or unjustified permissions. The low user count may reflect niche functionality rather than malicious activity, especially given the absence of supply chain indicators like typosquatting or recent publish dates. Overall, the evidence points to a legitimate tool with no security risks detected. The extension's publisher, MarkusvanKempen, is not flagged as a known malicious actor, and the extension's version (1.6.6) indicates active development. No findings of credential theft, data exfiltration, or supply chain poisoning were identified. The absence of obfuscation in non-locale files and lack of suspicious process spawning further confirm benign behavior. Users can safely install this extension for AI agent management without security concerns.
Key Reasons
- No security findings detected in evidence bundle
- Filesystem/process access aligns with legitimate AI agent management
- Low user count lacks corroborating malicious indicators
Reviewed 2026-04-23 · version 1.6.6; recommended action: no action; model confidence 90%.
VS Code version history
Risk trend by version
4 analyzed versions. Each point is the latest successful scan for that version; failed zero-score scans are hidden. Dates: First seen by Risky Plugins.
Pick any point on the chart, then open the Source tab to read that version's code.
Source Code Not Available
Source code is not available for this version of the extension.
About This Extension
Frequently Asked Questions
Similar Extensions
Related extensions from the same publisher or marketplace
WxO ToolBox(v2)
Markus van Kempen
WxO ToolBox
Markus van Kempen
IBM Maximo MCP for AI (Multi-IDE Connector & Tools)
Markus van Kempen
Maximo API Explorer — IBM Maximo REST API Browser, Tester & Scaffold Generator
Markus van Kempen
C2G Item-Passport
dazevichweb
CrudOnArray
KeshavSoft