acidrain-xss-security-testing
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill recommends downloading external scripts by cloning a repository from an untrusted GitHub account ('henry-lewiskpp1107/acidrain-security-script-hub.git').
- [DATA_EXPOSURE_AND_EXFILTRATION]: Contains utilities to extract browser cookies and exfiltrate them to a remote server. The file 'scripts/javascript/cookie-extractor.js' implements cookie harvesting and uses the fetch API to send the data to a remote URL.
- [DYNAMIC_EXECUTION]: Generates and executes code at runtime using dangerous sinks. The 'testBasicXSS' function in 'scripts/javascript/xss-basic.js' uses the 'innerHTML' sink to inject and execute script-based payloads.
- [INDIRECT_PROMPT_INJECTION]: Subject to indirect prompt injection through unsanitized processing of external URL parameters. Ingestion points: Processes URL parameters via 'window.location.search' in 'scripts/javascript/dom-xss-analyzer.js'. Boundary markers: None present; external data is processed directly. Capability inventory: DOM modification, network requests, and database query execution. Sanitization: Intentionally absent to facilitate vulnerability testing.
Audit Metadata