Teradata · dtehan-td · Jun 24, 2026 · Jun 23, 2026 · Jun 23, 2026 · Jun 23, 2026
diff --git a/.gitignore b/.gitignore
@@ -1,6 +1,8 @@
 __pycache__/
 logs/
 working/
+results/
+.mypy_cache/
 .env
 .DS_Store
 .vscode/
@@ -18,4 +20,7 @@ test_*.py
 .planning/
 .ruff_cache/
 Start_MCP_Server.bat
+.deepeval/
+.pytest_cache/
+
 
diff --git a/ddls_extracted/evals_employees_DDL.sql b/ddls_extracted/evals_employees_DDL.sql
diff --git a/ddls_extracted/evals_orders_DDL.sql b/ddls_extracted/evals_orders_DDL.sql
diff --git a/pyproject.toml b/pyproject.toml
@@ -24,13 +24,15 @@ classifiers = [
 ]
 dependencies = [
     "pydantic>=2.8.0,<3.0.0",
+    "pydantic-settings>=2.14.2",
     "fastmcp>=3.2.0,<3.4.2",
     "mcp[cli]>=1.0.0,<2.0.0",
     "teradatasqlalchemy>=20.0.0.0",
     "python-dotenv>=1.0.0",
     "PyYAML>=6.0.0",
     "sqlalchemy>=2.0.0,<3.0.0",
     "starlette>=1.3.1",
+    "langsmith>=0.8.18",
 ]
 
 [project.optional-dependencies]

diff --git a/scripts/seed_tdml_summaries.py b/scripts/seed_tdml_summaries.py
diff --git a/tests/eval_test/.deepeval/.deepeval_telemetry.txt b/tests/eval_test/.deepeval/.deepeval_telemetry.txt
@@ -0,0 +1,4 @@
+DEEPEVAL_ID=66f13728-e025-470d-b8fc-582a61a9d898
+DEEPEVAL_STATUS=old
+DEEPEVAL_LAST_FEATURE=evaluation
+DEEPEVAL_EVALUATION_STATUS=old
diff --git a/tests/eval_test/.env.example b/tests/eval_test/.env.example
@@ -0,0 +1,33 @@
+# AWS credentials — boto3 uses the standard credential chain:
+# env vars → ~/.aws/credentials → IAM instance profile
+AWS_REGION=us-east-1
+AWS_ACCESS_KEY_ID=your-access-key-id
+AWS_SECRET_ACCESS_KEY=your-secret-access-key
+# AWS_SESSION_TOKEN=your-session-token   # only needed for temporary credentials
+
+# Claude model ID on Bedrock — used for both agent and judge unless overridden
+BEDROCK_MODEL_ID=anthropic.claude-3-5-sonnet-20241022-v2:0
+
+# Override judge model separately (optional — defaults to BEDROCK_MODEL_ID)
+# BEDROCK_JUDGE_MODEL_ID=anthropic.claude-3-5-sonnet-20241022-v2:0
+
+# MCP server URL (must be running before evals start)
+MCP_SERVER_URL=http://127.0.0.1:8001/mcp
+
+# Database where evals_employees and evals_orders are created (usually your ClearScape username)
+EVALS_DATABASE=your_database_name
+
+# Teradata connection — used by setup_test_data.py and teardown_test_data.py only
+TERADATA_HOST=your-clearscape-host.teradata.com
+TERADATA_USER=your_username
+TERADATA_PASSWORD=your_password
+
+# Max tool-use turns per eval case (single-turn cases)
+AGENT_MAX_STEPS=5
+
+# Max tool-use turns per conversation turn (multi-turn cases)
+AGENT_MAX_STEPS_PER_TURN=3
+
+# Tool descriptions seen by the agent during evals (default: live MCP server)
+# USE_DESCRIPTION_OVERRIDES=1
+# DESCRIPTION_OVERRIDES_FILE=description_overrides.json