Command-line binary analysis with radare2/r2 — recon, disassembly, strings/imports, cross-references, patching, diffing, and scripting, no GUI needed. Load for exe/dll/so/elf/dex/wasm CLI analysis, quick triage before committing to IDA/Ghidra, r2 batch commands (-c/-A), r2pipe scripts, or help with
v2026.10.14-ee0e0c53 установки+3 за 7 дней★ 27 ЗвездыMITсинхронизировано 22 часа назад
You need fast, scriptable analysis of an in-scope binary (scope.txt / client-owned / your own sample —
tradecraft-scope-roe) from a terminal: first-pass recon before opening a heavyweight decompiler,
CI/SSH environments with no GUI, quick patches and diffs, or r2 automation. When you need
pseudocode-grade reading, hand off to reverse-eng-ida or reverse-eng-ghidra — r2's job is speed
and scriptability, not decompilation. For obfuscated JavaScript, that's reverse-eng-deobfuscation,
not this skill.
Every r2 operation is a composable command — the same one-liners work interactively, in batch
(-c "..."), over HTTP, or via r2pipe from any language — so recon that costs minutes in a GUI costs
seconds here, and every step is reproducible by construction.
Check the install, don't assume it:r2 -v / rabin2 -v. On Windows the binaries are
radare2.exe, rabin2.exe, rasm2.exe, radiff2.exe, rahash2.exe, rax2.exe, r2pm.exe.
Recon before analysis. Minimum first pass:text
rabin2 -I sample.exe # format, arch, bits, entry pointrabin2 -z sample.exe # stringsrabin2 -i sample.exe # imports (required — never skip to function-level work without it)rabin2 -E sample.exe # exports (required too for DLL/SYS)
Classify the imports (network / file / crypto / injection / registry) in your notes. An empty or
"too clean" import table means likely dynamic loading — plan to catch APIs at runtime instead of
concluding statically. No traditional IAT (.NET) → use IL/metadata inspection as the equivalent.
Packed sample: repair the IAT (ImportREC on x86, Scylla on x64); if repair fails, go dynamic rather
than grinding on a broken static view.
Interact only after recon.r2 sample.exe, then aaa (not the heavier aaaa), afl to list
functions, iz strings, s entry0 / pdf to disassemble a function, VV for graph view.
Locate the key logic:afl~main, iz~http, then axt <addr> to find who references a string
or address, s <addr> + pdf to read it. Full command reference: cheatsheet.md.
Patch only when asked, on a backup:r2 -w sample.exe (or oo+ in-session), then
wa nop / wa jmp 0x401050 / wx 9090, wq to write and quit. Warn before entering write mode;
re-disassemble to verify the change.
Batch and automate:r2 -A -q -c "afl;iz;ii;q" sample.exe for one-shot output; -A auto-
analyzes, -q quiet, -c runs a command string. Keep long command chains readable — split them.
Windows .sdb warnings (Cannot find ...\share\format\dll\*.sdb from rabin2) are usually
harmless if the main output is complete — don't mistake them for analysis failure.
Don't lead with aaaa — it's much heavier than aaa and rarely needed on the first pass.
Read-only by default — plain r2 <file> never modifies; only -w/oo+ does. Back up first.
Windows quirks — quote paths with spaces; if r2 isn't found after install, open a fresh
terminal (PATH refresh).
Ecosystem extras are accelerators, not replacements — r2pm -ci r2ghidra adds a decompiler,
r2http (r2 -N -e http.bind=localhost -e http.port=9393 -q -c=h sample.exe then
curl --data-binary 'aflj' http://127.0.0.1:9393/cmd) gives a stateful HTTP channel,
radius2 does symbolic execution. Same recon discipline applies.
A recon summary you can stand behind — format/arch/entry, the import classification, the suspicious
strings — plus located key functions (name + address + disassembly excerpt) and, for patches, a
re-disassembly showing the bytes changed as intended. Every step is a command someone else can rerun.