ghidra-headless-frida-runtime-injection

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents and generates command-line instructions for the Frida instrumentation tool (e.g., frida -f <target> -l <script>.js). This allows the agent to execute runtime analysis on binary targets using the provided local script library.\n- [PROMPT_INJECTION]: The planning-brief.md and SKILL.md files establish strict operational constraints, such as 'headless-only' and 'reproducible CLI capture'. These instructions are intended to enforce a specific technical workflow and prevent the agent from reverting to interactive or GUI-based modes, which are common failure points in automated analysis tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 03:31 PM
Security Audit — agent-trust-hub — ghidra-headless-frida-runtime-injection