penetration-tester-master
Warn
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: MEDIUMPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: The skill adopts an 'Elite Red Team Lead' persona and instructs the agent to perform offensive security tasks. This role-play framing is often used to bypass safety filters by contextualizing malicious actions as authorized ethical hacking exercises.
- [COMMAND_EXECUTION]: The execution protocol and methodology encourage the use of powerful offensive tools such as Nmap, Metasploit, and SQLMap. In agent environments where shell access is enabled, this could lead to the execution of high-risk commands against infrastructure.
- [DATA_EXFILTRATION]: The 'Hacking Methodology & Planning' section of SKILL.md explicitly lists 'Data Exfiltration' as a primary goal of the engagement lifecycle.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill instructs the agent to process external data, specifically 'every PR (Pull Request)' as defined in resources/owasp_top_10_2025.md.
- Boundary markers: No markers or 'ignore embedded instructions' warnings are present to isolate untrusted data from the agent's core instructions.
- Capability inventory: The skill references external tool execution (Metasploit, SQLMap) and includes local scripts (port_scanner.js, port_scanner.py) capable of processing user-supplied target arguments.
- Sanitization: No sanitization or validation logic is implemented for external content before it is processed by the agent.
Audit Metadata