Update default model to gpt-5.1

[devin-ai-integration]

Update default model to gpt-5.1

Summary

Changed the default OpenAI model from gpt-5 to gpt-5.1 in the Settings configuration. When users don't specify a MODEL environment variable, storymachine will now use gpt-5.1 instead of gpt-5.

Files changed:

• src/storymachine/config.py: Updated default model value
• tests/test_config.py: Updated test assertion to match new default

Review & Testing Checklist for Human

• Verify that gpt-5.1 is a valid and available OpenAI model name
• Run an end-to-end test without setting the MODEL environment variable to confirm the application works with the new default
• Consider if README or other documentation mentions the default model and needs updating

Test Plan

1. Unset the MODEL environment variable (or use a fresh environment)
2. Run uv run storymachine --prd <test-prd> --tech-spec <test-spec> --repo <test-repo>
3. Verify the application successfully uses gpt-5.1 and generates stories

Notes

• All existing tests pass with the updated default
• Users who have MODEL explicitly set in their environment are unaffected by this change
• The model detection logic in ai.py already supports gpt-5.x models via the model.startswith("gpt-5") check, so no changes were needed there

---

Requested by: nilenso ([email protected] / @tfidfwastaken)
Devin session: https://app.devin.ai/sessions/87db1c328d3e4221a705de21c27ba506

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring

[claude]

Code Review: Update default model to gpt-5.1

Summary

This PR updates the default OpenAI model from gpt-5 to gpt-5.1. The changes are minimal and focused, affecting only the configuration default and its corresponding test.

✅ Code Quality & Best Practices

Positive aspects:

• Minimal scope: The change is appropriately scoped to just the default value and its test
• Type safety: No type issues introduced - the field remains str
• Follows conventions: Adheres to CLAUDE.md guidelines (type hints present, simple focused change)
• Test coverage: The test was updated to match the new default value

Code structure:

• src/storymachine/config.py:10: Clean default value update
• tests/test_config.py:142: Test assertion correctly updated

⚠️ Concerns & Questions

1. Model Validity (CRITICAL)

The PR description includes a checklist item: "Verify that gpt-5.1 is a valid and available OpenAI model name."

Issue: As of January 2025 (my knowledge cutoff), OpenAI has not publicly announced a gpt-5.1 model. The standard naming convention for OpenAI models includes:

• GPT-4 series: gpt-4, gpt-4-turbo, gpt-4o
• O-series: o1, o1-preview, o1-mini, o3-mini

Recommendation:

• Verify this model name is valid in OpenAI's API before merging
• If this is a future/beta model, consider adding a comment explaining this
• Test against the actual OpenAI API to ensure it doesn't fail with "model not found"

2. Compatibility Check

The model detection logic in src/storymachine/ai.py:43 correctly handles gpt-5.x models via the prefix check:

or (model.startswith("gpt-5") and not model.startswith("gpt-5-chat"))

✅ This will work for gpt-5.1, so no compatibility issues there.

3. Documentation

The README.md doesn't explicitly mention the default model anywhere. While not required, users might benefit from knowing:

• What model is used by default
• How to override it with the MODEL environment variable

Optional improvement: Consider adding a brief note in README.md about model configuration.

🔍 Testing Recommendations

Per the PR description checklist, the following tests should be performed:

1. End-to-end test without MODEL env var:

   unset MODEL
   uv run storymachine --prd <test-prd> --tech-spec <test-spec> --repo <test-repo>

   This will verify the model actually exists and works with OpenAI's API.

2. Verify existing users are unaffected:
   Users with MODEL set in their environment should see no changes.

🔒 Security & Performance

• Security: No security concerns - this is just a configuration default
• Performance: Model choice affects API costs and response times
  • If gpt-5.1 is more expensive or slower than gpt-5, users should be aware
  • Consider documenting cost/performance implications

📋 Final Verdict

Status: ⚠️ Needs Verification Before Merge

Must address:

1. Confirm gpt-5.1 is a valid, publicly available OpenAI model
2. Test against OpenAI API to ensure it doesn't error

Nice to have:

1. Document the default model in README.md
2. Note any cost/performance differences from gpt-5

Once model validity is confirmed, the code changes themselves are clean and follow best practices. The implementation is correct and maintains backward compatibility.

---

Review based on CLAUDE.md guidelines and codebase conventions

[devin-ai-integration]

Closing due to inactivity for more than 7 days. Configure here.