Skills
Navigate
skills/wso2/financial-services-accelerator/set-up-wso2-open-banking/Gen Agent Trust Hub

set-up-wso2-open-banking

Fail

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches required dependencies from well-known and trusted sources. It downloads the MySQL JDBC driver from Maven Central, WSO2 product extensions from the official WSO2 documentation site, and CA certificates from Atlassian's Open Banking wiki.
  • [COMMAND_EXECUTION]: Runs shell scripts included within the WSO2 product and accelerator packages, such as merge.sh, configure.sh, and wso2server.sh, as part of the standard installation and startup process.
  • [PROMPT_INJECTION]: The skill accepts URLs from the user to download base product and accelerator zip files. It subsequently extracts these archives and executes scripts contained within them, creating an attack surface for indirect prompt injection if a malicious URL is provided.
  • Ingestion points: User-provided URLs for base product and accelerator zip files in SKILL.md (e.g., IS zip source, IAM accelerator zip source).
  • Boundary markers: None identified; instructions do not explicitly warn about ignoring embedded commands in user-provided archives.
  • Capability inventory: Execution of scripts such as merge.sh, configure.sh, wso2server.sh, and update_tool_setup.sh across product home directories.
  • Sanitization: None identified; the skill assumes the integrity of the contents of the downloaded archives provided by the user.
Recommendations
  • HIGH: Downloads and executes remote code from: https://repo1.maven.org/maven2/com/mysql/mysql-connector-j/9.6.0/mysql-connector-j-9.6.0.jar - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 16, 2026, 05:20 AM
Security Audit — agent-trust-hub — set-up-wso2-open-banking