apk-reverse/SKILL.md

name, description

name

description

apk-reverse

Reverse engineer, debloat, de-ad, patch, or re-sign Android APKs, and analyze their runtime and server-side behavior. Use when a task involves an .apk/.aab/.dex/.so sample, smali or dex patching, Frida/objection runtime hooking, repacking and re-signing, removing ads or SDK trackers, probing a mobile app's HTTP API, or deciding whether a client-side patch is even capable of achieving the goal. Covers recon, anti-tamper, ad removal, membership/paywall limits, dex-level surgical patching, repack pitfalls, device and emulator setup, and a hard-won failure catalogue.

APK Reverse Engineering & Patching

Goal of this skill: get to a verified, installable, still-working artifact fast, and avoid the whole class of mistakes that silently destroy an APK.

Three rules override everything else in this skill:

1. Deliver it in the environment the request actually requires. "It works" is not the goal; "it works under the stated constraint" is. A result obtained with root, or with live instrumentation, or with a host proxy, frequently does not satisfy a request for an installable artifact that works on a normal device — and it is easy to present that result as a finished one. Write the constraint down as a testable sentence in the first minutes and re-read it at every checkpoint. If you cannot meet it, say so and label any privileged workaround as a fallback, never as the deliverable. Full treatment: references/long-task-discipline.md §the most expensive drift.
2. Never ship an unverified APK. "It assembles" is not "it works". Install it, launch it, exercise the feature you changed, on a real device or a close emulator.
3. Know which layer owns the behavior before you patch. Ads, paywalls, and feature gates live in different places. Patching the wrong layer either does nothing or breaks the app. Classify first (see step 2 below), patch second.

Symptom index — look here before you debug from first principles

You arrived at a symptom, not at a file name. This table maps what you are seeing to the file that has already paid for that lesson. Skim it whenever something fails in a way the current plan does not explain; the cost of one lookup is minutes, and every entry below cost someone hours.

What you observe	Load first
A repackaged/re-signed build dies before your code runs; SIGSEGV, all registers zero, pc=0, fault addr near 0x0	native-tamper-and-suicide.md (deliberate crash), then code-virtualization-and-custom-linkers.md
No packer (Application is the app's own, dex readable) and it still dies	code-virtualization-and-custom-linkers.md
Log says a Java-layer signature/integrity check passed, yet the process dies	code-virtualization-and-custom-linkers.md §a Java-layer signature killer is a decoy
Deleting a library fixes validation but yields UnsatisfiedLinkError: dlopen failed: library "X" not found	code-virtualization-and-custom-linkers.md §the deadlock
Whole classes appear as bare native declarations with no body	code-virtualization-and-custom-linkers.md
A library's SONAME does not match its filename	code-virtualization-and-custom-linkers.md, native-and-so.md
Your edit had no effect at all, with no error	packers.md §map the validation boundary
Process hangs with no crash record, or dies to a uid 0 killer	native-tamper-and-suicide.md §the rule (you probably made a terminate path not return)
Death looks like an ordinary null dereference in a hardened library	native-tamper-and-suicide.md §deliberate-crash stubs
The app dies only while you are attached/rooted	detection-and-anti-analysis.md; run the unmodified original under identical conditions first
An install "succeeded" but nothing changed, or the version did not move	long-task-discipline.md §keep the observation window clean
Evidence contradicts itself, or a capture looks like two states mixed	long-task-discipline.md §keep the observation window clean
A script will not start, or a tool "is missing"	scripts/doctor.py, then toolchain.md
Feature-scoped network failure (login/register/pay) while the rest works	tls-and-cert.md — do not assume your patch caused it
Everything works but every signed request fails after repack	signature-derived-keys.md
Ads still appear after a patch that should have killed them	ad-removal.md §step 4 (count the SDK's own log lines; n -> 0, not "I did not see it")
A forced-update or "must update" gate blocks the build	updates-and-forced-upgrade.md §step 6
The dialog is gone but the feature is still locked	membership-and-limits.md / account-gates.md — decide server vs client authority before patching again
You are about to discard a route as "blocked"	packers.md — re-read it before writing any route off; mis-attributed failures have removed viable routes for hours
The task has run long and you are unsure what is already proven	long-task-discipline.md §keep a live record

Start here: classify the target in thirteen questions

Answer these before touching a tool. Every one of them changes the whole plan.

1. Is the app packed/hardened? → references/recon.md Read the manifest's application android:name. If it is a third-party shell class rather than the app's own Application, you have a packer and must handle it first.

2. Where does the behavior you want to change actually live?

   • Ad SDK (Pangle/GDT/AnyThink/Kuaishou/Baidu/Sigmob…) → usually client-side and removable → references/ad-removal.md
   • Server-issued ad config / sponsored cards → client renders server data → references/ad-removal.md §server-driven
   • Membership / VIP / paid content → usually server-authorized, client patch is cosmetic → references/membership-and-limits.md (read this before spending hours)
   • Feature flag, UI gate, debug switch → usually client-side
   • Anything decided by an API response → server-side → references/server-api.md

3. Is the app's own code in plain dex, or moved to native/Flutter/Unity? Plain dex → you can patch. Flutter (libflutter.so + libapp.so) / Unity (libil2cpp.so) / pure native → different toolchain entirely. See references/recon.md §code-location and references/framework-runtimes.md. Runtime check (cheap -- do it before committing to a layer): hook the obvious Java classes for the UI you care about, then reproduce that UI. If those hooks fire, the behavior is Java-owned. If they fire zero times while the UI is plainly on screen, the behavior is drawn by the runtime or by native code, and a dex-only plan will stall. Do not keep hunting in dex after a zero-hit probe -- that is the most expensive wrong turn in this skill's history.

4. What must the deliverable be able to do? Write the answer as a testable sentence before planning anything, then re-read it at every checkpoint. This is the drift guard, and the drift it guards against is the most expensive one in this skill: a runtime-only result (a data edit, a live hook, a blocked hostname, a host proxy) can look like success while failing the actual requirement. The axes that decide it: privilege (unrooted?), modification form (a rebuilt, installable artifact, or is live instrumentation acceptable?), ABI/device class, network (must it work online?), persistence (survives restart / upgrade / fresh install?), distribution (must the shipped file be self-contained?). → references/long-task-discipline.md §the most expensive drift.

5. Does the app verify its own signature, or does the server? App-side → you must bypass it. Server-side → re-signing silently breaks the app later. See references/repack-and-sign.md and references/server-api.md.

6. What is your device situation? → references/environment.md Rooted real device (best), emulator with root, or no device (static only). Also: this determines whether Frida is usable. Run scripts/preflight.py before your first experiment, and again whenever a failure surprises you — device state, a dead device server, a leftover proxy, and clock drift all masquerade as a broken patch (pitfalls.md P9).

7. Which architecture is actually executing? → references/native-and-so.md §Cross-architecture getprop reports what the device claims and primaryCpuAbi reports what the package manager chose — neither is what is running. Only the live mapping is ground truth (scripts/lib_map.py). If the library you meant to patch is not mapped, a translator is in play, or the ABI differs from your assumption, that changes the plan more than any patch will.

8. Is one specific feature failing at runtime — login, registration, payment, an API-backed screen — while the rest of the app works? → references/tls-and-cert.md. A feature-scoped network failure is very often a TLS/certificate problem on one code path, not a consequence of your patch. The app can even carry two independent trust chains, so "other requests work" proves nothing. Rule this out in minutes before hunting for a signature check.

9. Was the input a build you did not produce (a "cracked"/"modded" APK circulating online)? → references/third-party-builds.md. Audit it before adopting it: such builds are frequently re-protected (sometimes with more layers than the original) and may carry injected components or endpoints. Never use one as a patching workbench.

   Long-task rule: if this is likely to run long, open references/long-task-discipline.md now and keep its record updated as you go. Re-read the refuted-conclusions and dead-routes sections before starting any new experiment. Losing earlier findings is the most expensive failure in this skill, and it is entirely preventable.

10. Does the client sign its requests with its own signing certificate? → references/signature-derived-keys.md. Grep for toCharsString() / signatures[0] / getPackageInfo(..., 64) before the first repack. If that value feeds a native HMAC/DES routine, the rebuilt APK must hardcode the original certificate value at every read site, or every signed request fails while the app still launches and looks healthy. This is the single most expensive silent failure in a repack, and 15 minutes of grep prevents it.

11. Does the app die on its own after a while — with no Java stack trace, or with a native crash that looks like a bug? → references/native-tamper-and-suicide.md. A hardened library that decides the build is tampered rarely calls kill. It more often arranges a fault (load a small constant, use it as a pointer) so the death looks like an ordinary defect, and the system then reports it as an app "crash" or "abnormal" dialog. Two rules before you touch anything: enumerate which mechanism actually fires (the signal and the tombstone split them apart), and neutralise by returning, never by making it not return — a spinning stub freezes the process and produces a symptom that looks nothing like the cause.

12. Will this build still be usable in a week? → references/updates-and-forced-upgrade.md If the app has any version check, upgrade prompt, or self-update path, an unpatched build can be turned off remotely or replaced by the official package. This is one or two edits and it decides whether the work is durable — do it as part of the build, not as a follow-up. Also check for a hot-update / remote-config channel, which can restore behaviour you removed without any version change at all.

13. Does the request touch sign-in or phone binding — "no login required", "skip binding", "guest ok"? → references/account-gates.md. The whole difficulty here is separating a client-side gate (patchable) from an account-scoped resource (the screen is empty because the server has no account to answer for — not patchable). Classify first; and never fabricate a session to satisfy a gate, which produces a state worse than being signed out.

The workflow, end to end

1. Preflight, then Recon — scripts/doctor.py is the cheapest possible first command: it reports which toolchains and scripts can actually run here, and surfaces the environment facts that poison experiments (clock skew, leftover adb forward/proxy, a device-side frida process already running, a tool installed off-PATH). Then scripts/preflight.py before anything else if a device is involved (it takes seconds and prevents a whole class of false conclusions), then references/recon.md. Manifest, package name, version, ABI, dex count, packer, embedded SDKs, where the app's own code lives. Ten minutes here saves hours. If it is packed, unpack before anything else (references/recon.md §unpacking): you cannot patch code you cannot read, the encrypted payload lengths tell you which dumped dex is the original, and a memory dump must be de-duplicated by hash and structurally validated before any of it is trusted. If recon says there is no packer but a re-signed build still dies, you are in the layer references/code-virtualization-and-custom-linkers.md covers — do not proceed on the assumption that "no packer" means "editable". If the app already dies on its own — especially at a roughly constant time after launch, or with a native crash — locate the mechanism before planning any patch (references/native-tamper-and-suicide.md, scripts/native_crash.py). Record the observed time-to-death: it is the baseline every later attempt is measured against, and without it a surviving run cannot be told from a changed schedule.
2. Extract strings and endpoints — build a picture of the app's API surface and SDK inventory from the dex string tables. No decompiler needed for this, and it is fast. Scripts: scripts/dex_strings.py.
3. Trace to the owning class — find the class that wraps the behavior (the app almost always wraps third-party SDKs in one helper). Reverse-lookup instructions: references/dex-patching.md §finding-the-call-site.
4. Decide the patch layer — client SDK call / client rendering / client data consumption / server contract. See the table in references/ad-removal.md.
5. Patch surgically — references/dex-patching.md. Prefer dexlib2 method-level rewriting over whole-tree smali round-trip. Whole-tree round-trip damages R8-optimized dex in ways that only show up at runtime.
6. Repack and sign — references/repack-and-sign.md. Do not strip the whole META-INF/. This single mistake destroys otherwise-correct builds. 6b. Neutralise the update path — before you call the build done. If the app checks for updates at all, add the two-layer patch (references/updates-and-forced-upgrade.md): no-op the update routine's entry, and force the version comparison to its "no update" side. A build that can be switched off or replaced remotely is not a deliverable, and this costs minutes here versus a rebuild later. Do the same for any remote-config or hot-update channel that could restore the behaviour you removed. 6c. Handle account gates only after classifying them — if the request mentions sign-in or binding, apply references/account-gates.md and state plainly which guarded screens become usable and which stay empty because their content is account-scoped.
7. Verify on device — references/environment.md + references/verification.md. Check: launches, the changed behavior actually changed, nothing unrelated broke, and the app reaches its normal UI with no blocking dialog. First prove the artifact actually changed on the device -- a package manager reporting success does not prove an interposed confirmation was accepted (P18). Capture continuously for the first ~20 seconds after launch, and look at the captures — sampling gaps are how a blocking modal goes unseen (P20), and a burst of images that were never inspected is not evidence. If the accessibility tree is empty, the image is the primary evidence rather than a fallback.
8. Log what you learned — if a failure cost you more than thirty minutes, add it to references/pitfalls.md. That file is the most valuable artifact in this skill.

Non-negotiable constraints

• Read-only inputs. Keep the original APK/dex untouched; work on copies. Always keep a known-good baseline to diff against.
• One variable at a time. If you change two things and it breaks, you learn nothing. Build a control (same pipeline, zero patches) and compare.
• Verify structure after every dex edit. scripts/dex_classdiff.py must report zero differences in class set and access flags for classes you did not intend to change.
• Do not patch a method that is widely shared. Before patching any helper, count its callers (scripts/find_refs.py). A Long.valueOf wrapper with 30 callers is not an ad-specific hook.
• Do not make an API fail to suppress a UI element. A 404/400 on an endpoint that other features depend on takes the whole screen down with it. Suppress at the data-consumption or render layer instead.
• Neutralise a native terminate path by returning, never by making it not return. A stub, stub patch, or function entry replaced with a spin or a self-branch does not suppress the check — it freezes the caller, holding whatever lock it had, and unrelated threads wedge behind it. The symptom (a hang, an external kill, a restart loop) looks nothing like the cause, and there is no crash record to explain it. Return a benign value, and prefer success (0) over failure (-1). Never touch the ordinary-path symbols (pthread_exit, exit, abort, snprintf, closedir). references/native-tamper-and-suicide.md
• Look before you conclude — and look while you wait. Execute, capture, and inspect; do not drive and sleep blind. A screen that is actually looked at answers in one step what coordinate-guessing cannot answer in five: the layout moved, a different dialog is up, a countdown is frozen, the text on screen says exactly why. Where the thing you are waiting on is visible, a sample you can inspect beats a duration you hoped was right, and byte-identical samples mean nothing is going to change. scripts/snap.py; references/environment.md §look at the screen.
• Put a timeout on every command, and calibrate it from measurement. An unbounded call turns a stall into "the task stopped making progress", which is indistinguishable from slow work and costs hours silently. Time the operation once, record it, then derive the bound from it — that is what makes slow and hung distinguishable. A deadline that passes is a measurement, not a verdict. references/long-task-discipline.md §bound every wait.
• Every claim needs evidence. "Probably", "should be", "in theory" are not findings. Either you observed it, or you label it unverified.
  • "Done" means the user-visible outcome, not an internal signal. A blocking dialog still on screen means the task is not done, however many errors disappeared from the log. Absence of a log line is absence of evidence, never evidence of success.
  • Do not discard a route on compound evidence. If a failure followed two simultaneous changes, the attribution is a hypothesis, not a finding. Re-run it single-variable before writing the route off -- mis-attributed failures have removed viable approaches for a long time.
  • Prove the device changed before measuring. Install success describes the request, not the app on disk. Confirm the artifact actually advanced, or every following observation describes the previous build.
• Attribute a failure to the right layer before patching again. When something stops working after a rebuild, first check whether the unmodified original fails the same way on the same device and network. Feature-scoped network failures in particular are frequently the app's own TLS/certificate problem (references/tls-and-cert.md); chasing a signature check that does not exist burns hours.

Reference index

Load only what the current step needs.

File	Load when
references/recon.md	Starting any new sample; identifying packer, SDKs, code location, ABI
references/packers.md	The app is packed/hardened, or an edit makes it die before your code runs. Also load before discarding any route as "blocked by the shell"
references/code-virtualization-and-custom-linkers.md	No packer, dex is readable, and a re-signed build still dies — whole classes turned into native declarations, a private loader with a mismatched SONAME, an embedded self-decrypting payload, or a Java-layer "signature killer" that logs success while a native check kills you. Covers the keep-it/drop-it deadlock and the string-redirect escape
references/framework-runtimes.md	The UI is not native (Flutter / React Native / Unity / Cordova), or Java-layer hooks fire zero times while the UI clearly works
references/dart-aot.md	The logic lives in a Dart AOT snapshot (libapp.so): pinning the Dart version, building a matching decompiler, the object pool and reference indexing, register/boolean conventions, locating and patching Dart code
references/native-and-so.md	Patching in a .so, needing code to run before the app's own code, hand-built native payloads that crash inside the linker, or deciding which library/ABI is actually loaded and executing
references/native-tamper-and-suicide.md	The process dies on its own (no Java stack, or a native crash that looks like a bug); you are about to neutralise a kill/exit/abort path; or a hardened library's sections/function boundaries look wrong
references/detection-and-anti-analysis.md	The app fights back: it dies after you attach, refuses to run on your device, detects root/hook/debugger, or your dynamic tool simply does not work in this environment. Read the first section before escalating — the right answer is usually to switch to static, not to fight the detector
references/toolchain.md	Choosing or invoking tools, something is not installed, a tool's output smells wrong, or you need to know which tools exist only as a GUI
references/ad-removal.md	Task involves ads, trackers, sponsored cards, splash/interstitial/reward
references/updates-and-forced-upgrade.md	The patched build must keep working over time; the app has any version check, forced-upgrade dialog, self-update installer, or hot-update/resource channel. Load this for essentially every build you intend to ship.
references/account-gates.md	Task mentions "no login required", "don't force sign-in", "skip phone binding", "guest mode"; or a screen/feature is unreachable signed-out. Also load before promising that an account-scoped screen will show anything
references/membership-and-limits.md	Task involves VIP, subscription, paid content, unlock, "fully cracked"
references/server-api.md	The behavior is decided by a response, or you need to know if a patch can even matter
references/dex-patching.md	Any actual editing of dex/smali, choosing a patch layer, choosing a tool
references/patch-audit.md	Proving a patch landed, or that it is legal: length-vs-bytes comparison, the equal-length blind spot, verifier-level legality (move-result* adjacency), text-matching patch traps, and how to report a missing patch
references/repack-and-sign.md	Rebuilding, signing, installing, or a repacked app misbehaves
references/signature-derived-keys.md	The app reads signatures[0]/toCharsString(), or a rebuilt APK installs and runs but every signed request fails (sign/_p/uth empty or -1)
references/runtime-data.md	Local state matters: DataStore, SharedPreferences, SQLite, protobuf caches, tokens — or your data edit keeps being reverted, or a stored value looks encrypted
references/dynamic-frida.md	Frida setup, hooking strategy, tracing caller chains, finding the real call site
references/environment.md	Device/emulator setup, root, ADB, networking, offline devices, emulator console control and recovery, the preflight check to run before every experiment block
references/verification.md	Defining what "done" means; building the evidence chain
references/tls-and-cert.md	One feature fails at runtime (login, registration, payment, an API-backed screen) while the rest of the app works
references/third-party-builds.md	The input is a "cracked"/"modded" build you did not produce — audit it before trusting it
references/long-task-discipline.md	The task will run long, or you are resuming one. Live record, conclusion grading, drift checkpoints, deliverable-form drift (rooted-only vs shippable), bound-your-waits, handover
references/pitfalls.md	Always worth a skim before building. This is the failure catalogue.

Script index

All scripts are parameterized and path-agnostic; pass paths explicitly. Run --help or read the header of each.

Script	Purpose
scripts/doctor.py	Run this first. Capability report and per-script runnability: which tools exist (including ones installed off-PATH or as java -jar jars), which scripts can actually run here, and the environment facts that silently poison experiments — clock skew, leftover adb forward/proxy, a device-side frida process already running
scripts/smtool.py	baksmali/smali wrapper with a bundled classpath (assemble/disassemble dex)
scripts/patch_smali.py	Method-body replacement in a smali tree, matched by signature
scripts/dex_strpatch.py	Byte-level string constant patch with string_ids ordering guard
scripts/dex_classdiff.py	Compare two dex class tables (set + access flags) to prove a patch was surgical
scripts/dex_strings.py	Dump/extract strings and endpoints from dex without a decompiler
scripts/dart_pprefs.py	Build/query the object-pool offset -> code-site index for a Dart AOT snapshot (arithmetic decode; seconds, not minutes)
scripts/dart_pool_strings.py	Recover string literals from a Dart AOT snapshot: framed entries, the one-byte vs UTF-16 split, file offsets, and a run-length noise filter
scripts/dart_disasm.py	Annotated windowed disassembly of Dart AOT code (pool + boolean annotations) plus a B/BL caller index
scripts/find_refs.py	Count and list callers of a method/field (blast-radius check)
scripts/repack.py	Rebuild APK with replaced dex, strip only signatures, sign
scripts/dexpatch/	dexlib2 method-level surgical rewriter (the preferred patch tool) + build notes
scripts/devsh.py	Quoting-safe ADB shell helper for rooted devices
scripts/usb_net_proxy.py	Give an offline device network over USB (adb reverse + local proxy)
scripts/datastore_inject.py	Encode/inject AndroidX DataStore preferences (protobuf) safely
scripts/probe_api.py	Probe an app's HTTP API with correct headers, report status/shape
scripts/install_test.py	Install a build and run a launch/health check with logcat signal extraction
scripts/frida_probe.js	Four-layer runtime probe: app network layer + OkHttp + java.net + swallowed exception messages
scripts/run_probe.py	Inject a probe, stream it to a timestamped log file, stay resident while you operate the app
scripts/tls_check.py	Strict certificate check for one or more hosts (expired / wrong host / untrusted CA)
scripts/preflight.py	Read-only environment check before every experiment block: device, root, ABI/translation, clock skew, leftover proxy/forwards, dead device server. Run this before blaming a patch.
scripts/lib_map.py	What is actually mapped into a live process: per-library path, base, architecture (ELF e_machine), and classification (system / from-APK / runtime-materialized). Answers "which library and which ABI is really executing".
scripts/elf_plt.py	Resolve a PLT stub to its imported symbol on x86_64 and aarch64 (from the relocation table, not from position or a comment), list a symbol's callers, and byte-diff two libraries naming the symbol each changed stub belongs to. Run this before patching any stub, and to audit a patch set you inherited.
scripts/so_constpatch.py	Same-length in-place rewrite of an isolated string constant, for redirecting a library load instead of defeating a check (System.loadLibrary("checker") -> a library that is already mapped). Enforces equal length, refuses to touch a substring of a longer identifier, reports whether each hit sits in a constant pool, and patches inside an APK or a bare .so. See references/code-virtualization-and-custom-linkers.md.
scripts/apk_diff.py	Entry-level diff of two APKs: what changed, what was added (injection candidates), what was removed — by content hash, so same-size replacements are caught. Use it to audit a third-party build and to prove your own build was surgical.
scripts/native_crash.py	Locate a native death from a logcat capture or tombstone: signal, fault address, register state, backtrace split into your libraries vs system, the faulting instruction — plus an explicit flag when the fault looks arranged rather than accidental.
scripts/grab_crash.py	Recover a stack that a crash-reporter SDK swallowed, when the log shows the app died but prints no backtrace of its own.
scripts/blob_decode.py	Decode an opaque stored value by searching the parameter space (base64/base64url/hex × rotation × deflate/zlib/gzip) instead of guessing, then re-encode an edited payload with the same parameters.
scripts/snap.py	Bounded burst screenshot + control-tree capture, with a stall detector and an explicit verdict on whether the accessibility tree is usable at all. Use it so you look at the screen instead of driving blind.
scripts/sig_probe.py	Find the exact signatures[0].toCharsString() value: offline candidate enumeration from an APK (--apk), or the authoritative value read from a live package (--live). Feed the result into the hardcoded constant described in references/signature-derived-keys.md.