refactor: separate inline and context comment limits

Split REVIEW_MAX_COMMENTS into two independent variables:
- REVIEW_MAX_COMMENTS: total comments per PR (context review)
- REVIEW_MAX_INLINE_COMMENTS: comments per file (inline review)

Addresses AI reviewer feedback for independent configuration.

README.md

@@ -4,6 +4,7 @@
 A webhook-based service that automatically generates AI code reviews for Gitea Pull Requests using [xai-review](https://github.com/Nikita-Filonov/ai-review).
 
 ## Features
+*   **Integrated with AI**: Uses advanced LLMs for code analysis.
 *   **Zero-config per repository**: Works via a single Gitea System Webhook (or per-repo webhook).
 *   **Custom LLM Support**: Configuring it to use OpenAI, Ollama, vLLM, or any OpenAI-compatible endpoint.
 *   **Secure**: Runs in an isolated Docker container with strict context passed via environment variables.
@@ -24,7 +25,10 @@ Create a `.env` file with the following variables:
 | Variable | Required | Description | Default |
 | :--- | :--- | :--- | :--- |
 | `GITEA_TOKEN` | **Yes** | A Gitea Access Token. Required permissions: `repository` (Read), `issue` (Read and Write). | |
-| `GITEA_API_URL` | No | Full URL to your Gitea API (e.g., `http://gitea:3000/api/v1`). | |
+| `GITEA_API_URL` | **Yes** | Full URL to your Gitea API (e.g., `https://git.example.com/api/v1`). | |
+| `PORT` | No | Host port to expose the webhook server on. | `3000` |
+| `REVIEW_MAX_COMMENTS` | No | Maximum total AI comments per PR (context review). Set to `0` to disable. | `3` |
+| `REVIEW_MAX_INLINE_COMMENTS` | No | Maximum inline comments per file. Set to `0` to disable. | `3` |
 | `OPENAI_API_KEY` | **Yes** | API Key for your LLM provider. Use `dummy` for local models if no auth needed. | |
 | `OPENAI_BASE_URL` | No | Base URL for the LLM API. Set this for Ollama/vLLM (e.g. `http://host.docker.internal:11434/v1/`). | `https://api.openai.com/v1/` |
 | `LLM_MODEL` | No | The model name to request (e.g. `gpt-4o`, `llama3.1`). | `gpt-4o` |

compose.yaml

@@ -1,23 +1,27 @@
 
 services:
   ai-webhook:
-    build: .
+    image: gitea.ext.ben.io/b3nw/gitea-ai-webhook:latest
     container_name: gitea-ai-webhook
     restart: always
     ports:
-      - "3000:3000"
+      - "${PORT:-3000}:3000"
     environment:
       - GITEA_TOKEN=${GITEA_TOKEN}
       # LLM Configuration
       - LLM__PROVIDER=${LLM_PROVIDER:-OPENAI}
       - LLM__META__MODEL=${LLM_MODEL:-gpt-4o}
+      - LLM__META__TEMPERATURE=${LLM_TEMPERATURE:-0.2}
       # OpenAI / Compatible API Config
       - LLM__HTTP_CLIENT__API_TOKEN=${OPENAI_API_KEY}
       - LLM__HTTP_CLIENT__API_URL=${OPENAI_BASE_URL:-https://api.openai.com/v1/}
       # VCS Config
       - VCS__PROVIDER=GITEA
       - VCS__HTTP_CLIENT__API_TOKEN=${GITEA_TOKEN}
-      - VCS__HTTP_CLIENT__API_URL=${GITEA_API_URL:-http://gitea-server:3000/api/v1}
+      - VCS__HTTP_CLIENT__API_URL=${GITEA_API_URL}
+      # Review Config (inline=per file, context=per MR total)
+      - REVIEW__MAX_INLINE_COMMENTS=${REVIEW_MAX_INLINE_COMMENTS:-3}
+      - REVIEW__MAX_CONTEXT_COMMENTS=${REVIEW_MAX_COMMENTS:-3}
     volumes:
       - xai-cache:/root/.cache