agentleFS
jacksonmafra-umainjacksonmafra-umain/apk-lens/.claude/skills/analyze-apk/SKILL.md

analyze-apk

Skill0 starsChanged 2 months agoMITView on GitHub

Use when the user wants an Android app analysed — they give an APK/XAPK/APKS file path or a download link and ask what it collects, where it phones home, what permissions it wants, or whether it is safe. Runs the apk-lens pipeline and writes the report set.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. androguard/androguard★ 6.4ksame name
    2. TheQmaks/areclaw★ 67same name
    3. Slama-Consulting/proofbench★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository3 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Very specific:

    If jadx is missing, tell the user what it costs — no file:line evidence — give them the install command doctor printed, and offer to …

    line 25
  2. Distinctive:

    Never merge bundled-SDK call sites into the app's own behaviour.

    line 60
  3. Distinctive:

    No verdict about the developers.

    line 67
  4. Distinctive:

    Do not answer "is it spyware?" with yes or no.

    line 71
  5. Distinctive:

    Do not commit the APK or the report.

    line 74
  6. Distinctive:

    Do not skip the category, then present the permission verdicts as meaningful.

    line 75
  7. Distinctive:

    Do not paraphrase the coverage statement into something reassuring.

    line 76

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Analyse an Android app

When this applies

The user points at an Android app and asks what it does with data. Any of:

  • "Analyze this APK: ~/Downloads/app.xapk"
  • "What does this app collect?" with a file or link
  • "Where does this app phone home?"
  • "Is this app safe?" — answer the answerable part, and say which part is not

Do this

  1. Check the tooling (fast, and decides the depth):

    apk-lens doctor

    If jadx is missing, tell the user what it costs — no file:line evidence — give them the install command doctor printed, and offer to continue at --depth strings now.

  2. Pick a category. One of social messaging media camera navigation fitness shopping game utility browser. Infer it from the app; ask only if you truly cannot. It decides whether each permission reads as expected or as hard to justify — unknown produces verdicts worth nothing.
  3. Run it:

    apk-lens analyze "<PATH-OR-URL>" --category <CATEGORY> --depth full

    Tens of minutes on a large app at full depth. Say that before starting.

  4. Read 09-coverage-and-limits.md first, then README.md, then whichever of reports 02–07 the question is about. Use results.json for numbers.
  5. Write up, in this order:
    • what the app is built to do, in plain language
    • what stands out, worst first, every claim cited
    • what it notably does not do
    • what could not be seen — quote the coverage statement, do not soften it
    • what would settle the open questions
  6. Red-team yourself with apk-lens prompts 07 and report what changed.

The rules a finding is held to

  • Cite a file:line, an evidence/ file, or a results.json field. Otherwise delete the claim.
  • "The app contains a call to X" — never "the app sends X". Static analysis shows capability, not behaviour.
  • Never merge bundled-SDK call sites into the app's own behaviour.
  • Report absences: "does not request SMS", "no unattributed hosts". A report listing only what was found reads like an indictment.
  • Name the blind spots: native libraries read as text not disassembled, obfuscated identifiers, server-pushed config, cross-platform frameworks whose logic sits outside the decompiled Java.
  • Mark confidence: confirmed, likely, possible. Below that, drop it.
  • No verdict about the developers. Not legal advice.

Do not

  • Do not answer "is it spyware?" with yes or no. That question needs intent, and this method cannot establish intent. Say what the app is built to do, what is typical for its category, and what only a running test could settle.
  • Do not commit the APK or the report. Both are git-ignored by design.
  • Do not skip the category, then present the permission verdicts as meaningful.
  • Do not paraphrase the coverage statement into something reassuring.

Going deeper

apk-lens prompts lists focused prompts (permissions, endpoints, data collection, third parties, native/opacity, mitigation, plain-language summary). Each reads results.json rather than re-analysing the app.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.