[FEATURE] access the jadx's frida/xposed snippet

[viatm]

Feature Description

enable AI to access the jadx's frida snippet generation function when giving a poc or function analysis

Use Case

（Explain the use case or scenario where this feature would be valuable. How would you and others benefit from this feature?）

use case:When AI analyzes a component or vulnerability and asks AI to generate a POC, AI can access the Xposed or Frida snippet generation function to generate a more accurate POC.

benfit:
Frida can be used to dynamically modify the code, which can better locate the problem. A large number of Android reverse engineers will need this function. Some people may say that AI is enough to generate Frida scripts, which is true, but AI often cannot know the context of the function, while Jadx can handle this very well. In the latest version of Jadx, they have strengthened the processing of Frida snippets, which will bring better results.
xposed (lsposed) is also very popular. They are hook frameworks with a wide audience.

Proposed Solution

（If you have ideas about how to implement this feature, describe them here. Include any mockups, diagrams, or reference examples if applicable.）
Check the code of jadx to handle this. If it can handle the frida fragment, then the xposed fragment should be able to do it.

Alternatives Considered

(Have you considered any alternative solutions or workarounds? If so, please describe them.)

solutions: Or add some prompts to let ai automatically complete the context.

Additional Context

(Add any other context, references, or screenshots about the feature request here.)
At jadx,When you select a function and right click, you will see an option to copy the frida snippet
xposed snippet is also useful

[zinja-coder]

Thanks for the suggestion 🙏

This feature would actually be redundant, since the Jadx MCP server already provides the full application context (classes, methods, code, etc.) to the LLM. Once that information is available, you can simply prompt the LLM to generate the required Frida snippet directly based on the analysis results, without needing a dedicated feature for it.

So instead of building this into the server, the workflow is:

Use Jadx MCP to analyze and expose the app’s code.

Ask the LLM (in your prompt) to generate a Frida snippet for the desired hook/bypass/instrumentation using that context.

This keeps the server lightweight and avoids duplicating functionality that the LLM already handles well.