architecture.md

PR Review Processing Flow (Redis + Asynq + Worker)

flowchart LR
    ghwebhook[GitHub Webhook PR Open or Update] --> httpapi[HTTP Server API]
    httpapi --> redis[Redis Asynq DB]
    redis --> worker[Asynq Worker]
    worker --> ai[AI Review Engine LLM]
    ai --> ghapi[GitHub API Post Comments]

    httpapi --> validate[Validate Webhook]
    validate --> enqueue[Enqueue Task Asynq]
    enqueue --> redis

    worker --> dequeue[Dequeues Task]
    dequeue --> fetch[Fetch PR Diff]
    fetch --> buildprompt[Build Prompt]
    buildprompt --> callai[Call AI API]
    callai --> postcomment[Post Comments]
    postcomment --> ghapi

Loading

+----------------------+ +------------------------+ +------------------+ | GitHub Webhook | -----> | HTTP Server (API) | | AI Review | | (PR Open/Update) | | - Validates Webhook | | Generator (LLM)| +----------------------+ | - Enqueues Task (Asynq)| +------------------+ +------------------------+ | v +------------------+ | Redis (Asynq DB)| | - Stores Queued | | Tasks | +------------------+ | v +-----------------------------+ | Asynq Worker (Goroutine) | | - Dequeues Task | | - Fetches PR Diff | | - Builds Prompt | | - Calls AI API | | - Posts Comments via GitHub | +-----------------------------+

Understand graceful shutdown of services

flowchart TD
    main[main function] --> init[Initialize Dependencies]
    init --> sig[SIGINT Listener Goroutine]
    init --> runworker[Run Worker Goroutine]
    init --> startserver[Start Server Goroutine]
    sig --> wait[Wait for Shutdown]
    runworker --> wait
    startserver --> wait

Loading

┌──────────────┐ │ main() │ └────┬─────────┘ │ ├─ initializeDependencies() ├─ signal listener goroutine ───┐ ├─ RunWorker() goroutine │ (listens for ctx cancel) ├─ startServer() goroutine │ (listens for ctx cancel) └─ wg.Wait() <──────────────────┘ waits for both