[BUG] Remove hardcoded llama3.2 model, allow configurable Ollama model selection

[sujay-d07]

What happened?

The Ollama model is hardcoded to llama3.2 in multiple places (ask.py and tests), preventing users from specifying alternative Ollama models like llama3.1:8b, mistral, phi3, or other compatible models.

How to reproduce

1. Set environment variable: export OLLAMA_MODEL=mistral
2. Run: cortex ask "what is nginx?"
3. Expected: Cortex uses the mistral model specified in OLLAMA_MODEL
4. Actually happened: Cortex ignores the environment variable and defaults to llama3.2 hardcoded in ask.py

Error output

No error is thrown, but the hardcoded model is used instead of respecting configuration:

Current hardcoded locations:

ask.py - _default_model() returns hardcoded "llama3.2"
test_ask.py - Test expects hardcoded model
test_ollama_integration.py - Multiple test instances with hardcoded ollama_model="llama3.2"

Expected behavior

Read from OLLAMA_MODEL environment variable or config file, allowing users to specify any Ollama model.

[sujay-d07]

working on this issue - effiti

[coderabbitai]

📝 CodeRabbit Plan Mode

Generate an implementation plan and prompts that you can use with your favorite coding agent.

• Create Plan

Examples

• Example 1
• Example 2

---

🔗 Similar Issues

Related Issues

• Bug: Cortex attempts Ollama connection when not configured #316
• [FEATURE] Offline Mode Requires Manual Export of CORTEX_PROVIDER=ollama #371

👤 Suggested Assignees

• mikejmorgan-ai
• Suyashd999

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!