ADFA-4589 | Add suite of AI-powered developer plugins for Code on the Go by jatezzz · Pull Request #41 · appdevforall/plugin-examples

jatezzz · 2026-07-07T23:09:15Z

Description

Adds three new AI plugins and hardens the existing ai-core / ai-assistant (both already on main):

Code Suggestions - inline ghost-text completions as you type.
Speech-to-Text - voice input transcribed into the editor.
Vector Search - embeddings-based semantic search over the project (uses ai-core's getEmbeddings(), with a lexical fallback).

All three share ai-core's LlmInferenceService (on-device llama.cpp and/or Gemini) instead of bundling their own models.

Details

Stability & fixes to the existing plugins:

ai-core: reject embedding models on the chat path before they SIGABRT the IDE (ADFA-4388); load the user-selected model instead of the newest .gguf in Downloads; shut down the Llm-RunLoop executor on dispose to stop the classloader leak; move Gemini API key to the x-goog-api-key header.
ai-assistant: fetch the live Gemini model catalog (default gemini-2.5-flash, retire gemini-1.5-flash); fix settings-dialog UX and backend selection/routing.
Cross-cutting: cancel coroutine scopes on disposal; version ranges, READMEs, in-IDE help HTML, and day/night PNG icons across the AI plugins.

document_5073463760878832078.mp4

Ticket

ADFA-4589

Observation

llama.cpp is a submodule at ai-core/subprojects/llama.cpp (private fork, branch androidide-custom, needs SSH): git submodule update --init ai-core/subprojects/llama.cpp. A prebuilt ai-core/libs/v8/llama-v8-release.aar is committed, so building llama.cpp from source isn't required just to compile.
libs/*.jar are CI-managed and intentionally not committed here, but vector-search needs LlmInferenceService.getEmbeddings(), so the refreshed plugin-api.jar must be in place (CI / scripts/update-libs.sh) before building.

Notes:

Dropped the original .cgp / internal-docs observation: I verified none are tracked in this diff.
Corrected the submodule path (ai-core/…, not ai-assistant/…) and the "adds four plugins" claim (ai-assistant/ai-core already exist on main; this adds three).

claude

Claude Code Review

This repository is configured for manual code reviews. Comment @claude review for a one-time review, or @claude review always to subscribe this PR to a review on every future push.

_{Tip: disable this comment in your organization's Code Review settings.}

hal-eisen-adfa · 2026-07-11T02:33:12Z

Might be too early to properly review this. They all need version ranges, html documentation, tier 1/2/3 help, and README.md.

CoroutineScope(Dispatchers.IO) and never gets cancel() in dispose().

@volatile static pluginContext never cleared, native model never unloaded, llama executor never shut down.

ai-assistant has security problems

Gemini API key in plaintext SharedPreferences with a misleading "encrypted" comment (AgentSettingsStore.kt:39-40, AiAssistantPlugin.kt:176)
API key in URL query string (GeminiBackend.kt:336-341) — should be x-goog-api-key header
ai-core makes HTTP calls but doesn't declare network.access/INTERNET.

We also need the new plugin api jar otherwise these won't build.

Inline ghost-text AI code completions; restored onto main's split layout.

Voice-to-code speech input; restored onto main's split layout.

Semantic code search via embeddings; restored onto main's split layout.

… Downloads Stream the selected content:// URI into a private cache file; drop the Downloads fallback that silently loaded the wrong model (ADFA-4388).

…shing Add GgufModelInspector + UserFeedback; classify the .gguf and throw IncompatibleModelException before native decode SIGABRTs the IDE (ADFA-4388).

….5-flash listModels now queries the live v1beta ListModels catalog (generateContent-only); default bumped to gemini-2.5-flash and retired models migrated on load.

Readable settings title; real model name shown and persisted across reopen; and the chat backend indicator + routing now follow the selected backend, refreshing when the settings dialog closes.

…ggestion-plugin and vector-search-plugin names

- Add version ranges, README/docs, and tiered help coverage for AI plugins - Cancel plugin coroutine scopes during disposal - Move Gemini listModels API key from URL query to x-goog-api-key header - Document Gemini API key storage tradeoff and remove misleading encryption note - Update AI plugin review status with verification results

jatezzz · 2026-07-16T18:48:56Z

@hal-eisen-adfa Thanks for the feedback, I've addressed all the concers, attached the proof that the new plugins build with the new Code On the Go plugin api.

https://github.com/appdevforall/plugin-examples/actions/runs/29523333464

hal-eisen-adfa · 2026-07-16T20:36:51Z

Almost everything is now fixed.

One big gap that wasn't addressed yet: LLamaAndroid needs to expose a shutdown that calls (runLoop as ExecutorCoroutineDispatcher).close() / executor.shutdown(), invoked from AiCorePlugin.dispose().

LLamaAndroid's single-thread run-loop executor was never shut down, so the non-daemon Llm-RunLoop thread outlived the plugin and pinned its classloader. Expose LLamaAndroid.shutdown() (closes the ExecutorCoroutineDispatcher) and call it from LocalLlmBackend.close() after the model unload completes. Regenerate the prebuilt llama AAR so the new API is visible to the plugin build.

jatezzz · 2026-07-17T14:24:37Z

Added day and night icons for AI plugins

… light/dark theme - Agent tab now shows an error bubble + Snackbar ("Open Settings...") when no local model or Gemini key is configured, instead of silently doing nothing. - Theme UI to the IDE day/night setting: route inflation through PluginFragmentHelper.getPluginInflater, make PluginTheme a Material3 DayNight theme, add values-night/ colors, and replace hardcoded hex with @color refs. - Fix AI Settings toolbar contrast (colorOnPrimary title/icon on the primary bar).

jatezzz · 2026-07-21T22:47:19Z

Update theme and model validation on fresh installs

document_5114081575834749100.mp4

claude Bot reviewed Jul 7, 2026

View reviewed changes

jatezzz requested review from a team, Daniel-ADFA and itsaky-adfa July 7, 2026 23:09

jatezzz force-pushed the refactor/ADFA-4589-ai-hackathon-plugin branch from e1a6598 to 4986735 Compare July 7, 2026 23:37

jatezzz force-pushed the refactor/ADFA-4589-ai-hackathon-plugin branch from fc375e7 to d022464 Compare July 15, 2026 15:14

jatezzz added 9 commits July 15, 2026 16:33

feat: add code-suggestions plugin

ce03159

Inline ghost-text AI code completions; restored onto main's split layout.

feat: add speech-to-text plugin

8c51bef

Voice-to-code speech input; restored onto main's split layout.

feat: add vector-search plugin

8d39b42

Semantic code search via embeddings; restored onto main's split layout.

fix(ai-core): load the user-selected model instead of newest .gguf in…

990e180

… Downloads Stream the selected content:// URI into a private cache file; drop the Downloads fallback that silently loaded the wrong model (ADFA-4388).

fix(ai-core): reject embedding models on the chat path instead of cra…

b67f58a

…shing Add GgufModelInspector + UserFeedback; classify the .gguf and throw IncompatibleModelException before native decode SIGABRTs the IDE (ADFA-4388).

fix(ai-assistant): fetch live Gemini models and drop retired gemini-1…

a56ced5

….5-flash listModels now queries the live v1beta ListModels catalog (generateContent-only); default bumped to gemini-2.5-flash and retired models migrated on load.

fix(ai-assistant): settings dialog UX and backend selection bugs

40c4f60

Readable settings title; real model name shown and persisted across reopen; and the chat backend indicator + routing now follow the selected backend, refreshing when the settings dialog closes.

fix(code-suggestions-plugin): LLM service lookup

3446ad4

feat(vector-search): wire semantic project search, also fixed code-su…

9f7c87e

…ggestion-plugin and vector-search-plugin names

jatezzz force-pushed the refactor/ADFA-4589-ai-hackathon-plugin branch from c9ce9b8 to 9f7c87e Compare July 15, 2026 21:33

jatezzz and others added 2 commits July 16, 2026 11:09

Merge branch 'main' into refactor/ADFA-4589-ai-hackathon-plugin

9d6df06

hal-eisen-adfa approved these changes Jul 16, 2026

View reviewed changes

refactor: add icons day and night for AI plugins

239d443

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

ADFA-4589 | Add suite of AI-powered developer plugins for Code on the Go#41

ADFA-4589 | Add suite of AI-powered developer plugins for Code on the Go#41
jatezzz wants to merge 14 commits into
mainfrom
refactor/ADFA-4589-ai-hackathon-plugin

jatezzz commented Jul 7, 2026 •

edited

Loading

Uh oh!

claude Bot left a comment

Uh oh!

hal-eisen-adfa commented Jul 11, 2026

Uh oh!

jatezzz commented Jul 16, 2026

Uh oh!

hal-eisen-adfa commented Jul 16, 2026

Uh oh!

jatezzz commented Jul 17, 2026

Uh oh!

jatezzz commented Jul 21, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Uh oh!

Conversation

jatezzz commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Details

Ticket

Observation

Uh oh!

claude Bot left a comment

Choose a reason for hiding this comment

Claude Code Review

Uh oh!

hal-eisen-adfa commented Jul 11, 2026

Uh oh!

jatezzz commented Jul 16, 2026

Uh oh!

hal-eisen-adfa commented Jul 16, 2026

Uh oh!

jatezzz commented Jul 17, 2026

Uh oh!

jatezzz commented Jul 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

jatezzz commented Jul 7, 2026 •

edited

Loading

jatezzz commented Jul 21, 2026 •

edited

Loading