Fix deployment: catch syntax errors at build, clean Render config, and update OpenAI usage

[Nsethi16]

Motivation

• A deployment failure was caused by unresolved merge-conflict markers in app.py which produced a runtime IndentationError at startup.
• The Render service config duplicated the name key and used an outdated default model, which could cause confusion during blueprint deploys.
• The repository should fail fast at build time for Python syntax errors so malformed files are caught before runtime.

Description

• Removed/normalized conflicting lines in app.py, clarified agent role to a math tutor, and set the default OPENAI_MODEL to gpt-4.1 while configuring the Responses API code_interpreter tool with container: {"type": "auto", "memory_limit": "4g"}.
• Updated render.yaml to remove the duplicate service name, changed the default OPENAI_MODEL to gpt-4.1, and extended buildCommand to run python -m py_compile app.py after installing dependencies.
• Bumped openai SDK in requirements.txt to openai==1.55.3 and updated README.md to reflect the math-tutor focus, the model default, and the use of the code_interpreter tool.

Testing

• Ran python -m py_compile app.py and it succeeded, confirming no syntax errors.
• Performed Flask test-client sanity checks which passed: GET / returned 200 and POST /ask with an empty body returned 400.
• No automated test failures were observed for the modified code paths during local validation.

---

Codex Task

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.