Maat: Symbolic execution made easy

Summary: Trail of Bits released Maat, a cross-architecture symbolic execution framework built on Ghidra's p-code intermediate representation, with both C++ and Python APIs. The tool offers dynamic symbolic execution, taint analysis, binary instrumentation, and constraint solving, and has been tested only on X86 and X64 so far, with early measurements of 100,000 to 300,000 instructions symbolically executed per second on a 2.3 GHz Intel Core i7 laptop with 32 GB RAM.
What's compelling: Symbolic execution has long carried runtime overhead that kept it out of day-to-day workflows, and Maat's default settings prioritize soundness over performance, so analysts gain a lower-level toolkit whose speed depends heavily on how many SMT solver calls and analysis callbacks they add.

Read the original at Trail of Bits Blog ↗