Repository navigation
Conversation
added 4 commits
October 21, 2026 17:50
…roller scaffolding Introduce the core building blocks for an embedded llama-server (OpenCL) local AI experience: - LocalAiModel/LocalAiModelScanner discovers .gguf models on device storage - LocalAiProotCommand builds the sandbox prefix binding /vendor and /apex so the bionic-linked llama-server can reach the Adreno OpenCL driver and linker - LocalAiInstaller unpacks the llama-oc payload into the rootfs - LocalAiController owns the process lifecycle (start/stop/health) and exposes a state flow for the settings UI Part of yuga-hashimoto#338 (local AI on device).
…running - unpack llama-oc.tar (aapt2 gunzips the asset, no .gz name in apk) - bind model parentFile into proot instead of the file itself - add PROOT_TMP_DIR to the embedded proot environment - add SAF multi-file import UI with empty-state and result labels - scan extra external files dir for imported models
Local AI scaffolding never meant to remove the runtime/provider bindings (VoskModelStore, LocalRuntimeInstaller, LocalRuntimeProcessLauncher, ...); they are still referenced by AndCodeApplication and the runtime packages. Also ignore the local proot/llama asset and local.properties backup.
aixtin
changed the title
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
feat(local-ai): on-device local AI inference via llama.cpp (OpenCL)
Summary
Adds an on-device local AI option to and-code, powered by llama.cpp with the Adreno OpenCL backend. Users can import GGUF models, run llama-server in a proot sandbox, and chat with the local model directly from the app — no cloud dependency.
Verification (real device: OnePlus Ace 3 Pro / PJX110 / Adreno 750 / KSU)
Changes (4 commits)
Known limitations