skills/mukul975/anthropic-cybersecurity-skills/exploiting-sql-injection-vulnerabilities/Gen Agent Trust Hub
exploiting-sql-injection-vulnerabilities
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted HTTP responses from target web applications in 'scripts/agent.py', creating an attack surface where malicious data could influence agent behavior.
- Ingestion points: Data is ingested from remote servers in 'scripts/agent.py' using the 'requests' library across multiple detection functions including 'detect_error_based', 'detect_boolean_based', and 'detect_time_based'.
- Boundary markers: The script does not implement delimiters or 'ignore embedded instructions' warnings when processing or reporting remote content to the agent.
- Capability inventory: The script has the capability to perform network requests ('requests.get', 'requests.post') and write findings to the local file system ('json.dump').
- Sanitization: There is no sanitization, escaping, or validation of the remote content before it is included in the JSON report or processed by the agent.
Audit Metadata