sql-injection-union

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests and processes untrusted data from external targets and engagement state. Ingestion points: Data retrieved from target HTTP responses (Steps 3-5) and the get_state_summary() state management function in SKILL.md. Boundary markers: Absent; no delimiters are used to isolate untrusted data. Capability inventory: Command execution via sqlmap and burpsuite tools; file system write access for evidence logging in the engagement/evidence/ directory. Sanitization: Absent; the skill lacks mechanisms to validate or sanitize external input before use.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute complex command-line operations using sqlmap for automated exploitation.
  • [DATA_EXFILTRATION]: The skill is designed to facilitate the extraction of sensitive database information through UNION-based SQL injection techniques across various platforms.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 09:15 AM
Security Audit — agent-trust-hub — sql-injection-union