skills/mukul975/anthropic-cybersecurity-skills/performing-fuzzing-with-aflplusplus/Gen Agent Trust Hub
performing-fuzzing-with-aflplusplus
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill script scripts/agent.py interacts with the AFL++ toolchain via subprocess.run. Command arguments are constructed from user input and passed as lists, which prevents shell injection vulnerabilities.
- [EXTERNAL_DOWNLOADS]: References to external code point to the official AFL++ repository on GitHub. This is a well-known and trusted source within the security community.
- [COMMAND_EXECUTION]: The agent executes compilation and fuzzing tools based on user-provided paths. This behavior is consistent with the skill's primary purpose of automating security testing workflows.
- [SAFE]: No obfuscation, hardcoded secrets, or suspicious exfiltration patterns were identified during the analysis of the prompt or script files.
Audit Metadata