Merge branch 'feature/nada-rag' of https://github.com/NadaBhm/devguard-ai into feature/nada-rag

Author: NadaBhm

README.md

@@ -98,7 +98,7 @@ devguard-ai/
 │   │       ├── embeddings.py
 │   │       └── tests/
 │   │
-│   ├── subgroup2/               # Oussema + Hbib
+│   ├── agents/               # Oussema + Hbib
 │   │   ├── deployops/           # Oussema owns
 │   │   │   ├── agent.py
 │   │   │   └── tests/

requirements.txt

@@ -38,12 +38,4 @@ torch>=2.1.0
 numpy>=1.26.0
 
 # ========== Utils ==========
-python-dotenv>=1.0.0
-httpx>=0.27.0
-
-# ========== Testing ==========
-pytest>=8.0
-pytest-cov>=5.0
-pytest-asyncio>=0.23
-ruff>=0.4
-mypy>=1.10
+python-dotenv>=1.0.0

src/agents/agentInfraCost/.env.example

@@ -0,0 +1,6 @@
+# Optional. Enables real Gemini-powered explanatory text in llm_enrichment.py
+# (architecture_explanation / cost_summary / finops_justification).
+# If unset, the agent automatically falls back to deterministic template
+# text built from the already-computed decision/cost/finops values — the
+# pipeline runs end-to-end either way, with no blocking and no error.
+GEMINI_API_KEY=