Merge pull request #8 from serkanh/imporvement/model-selection-via-env-var

Imporvement/model selection via env var

Author: serkanh

.github/workflows/ci.yml

@@ -61,63 +61,18 @@ jobs:
         PYTHONPATH=$PWD pytest tests/ -v -k "not integration" --tb=short --maxfail=3
       continue-on-error: true
 
-    - name: Create test environment files
+    - name: Run full test suite with Docker Compose
       run: |
-        # Create agents/.env for CI testing
-        cat > agents/.env << EOF
-        # AI Model API Keys (dummy values for CI)
-        GOOGLE_API_KEY=dummy_key_for_testing
-        ANTHROPIC_API_KEY=dummy_key_for_testing
-        DB_PASSWORD=postgres
-        AWS_PROFILE=default
-        AWS_REGION=us-east-1
-        LOG_LEVEL=INFO
-        EOF
-
-        # Create slack_bot/.env for CI testing
-        cat > slack_bot/.env << EOF
-        # Slack App Configuration (dummy values for CI)
-        SLACK_BOT_TOKEN=xoxb-dummy-token-for-ci
-        SLACK_SIGNING_SECRET=dummy-signing-secret
-        SLACK_APP_TOKEN=xapp-dummy-app-token
-        SRE_AGENT_API_TIMEOUT=30
-        SESSION_TIMEOUT_MINUTES=60
-        MAX_ACTIVE_SESSIONS=100
-        HEALTH_CHECK_INTERVAL=300
-        EOF
-
-    - name: Start PostgreSQL
-      run: |
-        docker compose up -d postgres
-        sleep 10  # Wait for postgres to be ready
-
-    - name: Build test image
-      run: |
-        docker build \
-          --target test \
-          -f ./agents/sre_agent/Dockerfile-agent \
-          -t sre-bot:test \
-          .
-
-    - name: Run full test suite in Docker
-      run: |
-        docker run --rm \
-          --network sre-bot_default \
-          -e PYTHONPATH=/app \
-          -e DB_HOST=postgres \
-          -e DB_PORT=5432 \
-          -e DB_NAME=srebot \
-          -e DB_USER=postgres \
-          -e DB_PASSWORD=postgres \
-          sre-bot:test \
-          pytest tests/ -v --tb=short --maxfail=5
+        # Use dedicated CI docker-compose configuration
+        docker compose -f docker-compose.ci.yml up --build --abort-on-container-exit --exit-code-from sre-bot-test
 
     - name: Show test logs on failure
       if: failure()
       run: |
         echo "=== Test execution failed. Checking container logs ==="
-        docker compose logs postgres || echo "No postgres logs available"
+        docker compose -f docker-compose.ci.yml logs postgres || echo "No postgres logs available"
+        docker compose -f docker-compose.ci.yml logs sre-bot-test || echo "No test logs available"
 
     - name: Clean up test environment
       if: always()
-      run: docker compose down -v
+      run: docker compose -f docker-compose.ci.yml down -v

INITIAL.md

@@ -1,16 +1,42 @@
 ## FEATURE
 
-We need to setup authentication to accesss single or multiple aws systems so we can run queries. It will be role based access that will can assume a remote role that is pre-created and will be able to assume the role to run aws api queries. This is will be used for other
-agents as well so should be shareable.
+You are an experienced Python developer and you are given a task to implement a feature. We ned to implement a feature where will determine how to initate the agents/agent.py with the proper configuration.
 
 ## EXAMPLES
 
-[Provide and explain examples that you have in the `docs/examples/` folder]
+ If a user is sets env var Google Gemini with `GOOGLE_API_KEY` the initiation should be like this with GOOGLE_AI_MODEL env var that is set default to 'gemini-2.0-flash'
+ ```python
+ 
+ agent = Agent(
+        name="sre_agent",
+        model='gemini-2.0-flash',
+        instruction="""You are an expert Site Reliability Engineer (SRE) assistant specializing in operational tasks,
+        infrastructure management, and cost optimization.
+```
+ 
+If env var is set to Anthropic with `ANTHROPIC_API_KEY` the initiation should be like this with ANTHROPIC_MODEL env var that is set default to 'claude-3-7-sonnet-20240620'
+```
+agent = Agent(
+        name="sre_agent",
+        model=LiteLlm('claude-3-5-sonnet-20240620'),
+        instruction="""You are an expert Site Reliability Engineer (SRE) assistant specializing in operational tasks,
+        infrastructure management, and cost optimization.
+```
+
+if env var is "BEDROCK_INFERENCE_PROFILE" the initiation should be like this:
+
+```python
+agent = Agent(
+        name="sre_agent",
+        model=liteLlm('arn:aws:bedrock:us-west-2:812201244513:inference-profile/us.anthropic.claude-opus-4-1-20250805-v1:0'),
+        instruction="""You are an expert Site Reliability Engineer (SRE) assistant specializing in operational tasks,
+        infrastructure management, and cost optimization.
+```
 
 ## DOCUMENTATION
 
-[List out any documentation (web pages, sources for an MCP server like Crawl4AI RAG, etc.) that will need to be referenced during development]
+https://google.github.io/adk-docs/agents/models/#method-c-service-account-for-production-automation
 
 ## OTHER CONSIDERATIONS
 
-- Should be simple to configure perhaps requiring simply aws role arn and account id.
+