Update default model to OpenAI's GPT-5.5

[PeterDaveHello]

GitHub Copilot Pull Request summary

This pull request updates the default language model version used by PR-Agent to the latest available, ensuring improved capabilities and alignment with the latest model offerings. It also updates the documentation to reflect this change.

Model version updates:

• Updated the default model in configuration.toml from gpt-5.4-2026-03-05 to gpt-5.5-2026-04-23 to use the latest language model by default.

Documentation updates:

• Changed the installation instructions in locally.md to require access to GPT-5.5 (instead of GPT-5.4) and gpt-5.4-mini, ensuring users acquire the correct API key for the updated model.

[qodo-free-for-open-source-projects]

Review Summary by Qodo

Update default model to OpenAI's GPT-5.5

✨ Enhancement

Walkthroughs

Description

• Update default model from GPT-5.4 to GPT-5.5
• Align configuration with latest OpenAI model
• Update installation documentation for new model

Diagram

flowchart LR
  A["GPT-5.4-2026-03-05"] -- "upgrade to" --> B["GPT-5.5-2026-04-23"]
  C["configuration.toml"] -- "updated" --> B
  D["locally.md"] -- "updated" --> E["GPT-5.5 requirement"]

Loading

File Changes

1. pr_agent/settings/configuration.toml ⚙️ Configuration changes +1/-1

Update default model configuration

• Updated default model from gpt-5.4-2026-03-05 to gpt-5.5-2026-04-23
• Maintains fallback model as gpt-5.4-mini

pr_agent/settings/configuration.toml

---

2. docs/docs/installation/locally.md 📝 Documentation +1/-1

Update installation documentation

• Updated OpenAI API key requirement from GPT-5.4 to GPT-5.5
• Maintains requirement for gpt-5.4-mini access

docs/docs/installation/locally.md

---

[qodo-free-for-open-source-projects]

Code Review by Qodo

🐞 Bugs (1) 📘 Rule violations (0)

1. Outdated model comments 🐞 Bug ⚙ Maintainability

Description

After switching the default model to gpt-5.5-2026-04-23, the commented example settings for
model_reasoning/model_weak still reference gpt-5.4 variants, which can lead users to unintentionally
pin older models when they uncomment the examples.

Code

pr_agent/settings/configuration.toml[R7-10]

+model="gpt-5.5-2026-04-23"
fallback_models=["gpt-5.4-mini"]
#model_reasoning="gpt-5.4-mini" # dedicated reasoning model for self-reflection
#model_weak="gpt-5.4-nano" # optional, a weaker model to use for some easier tasks

Evidence

The default model is now a GPT-5.5 version, while the example commented model_reasoning/model_weak
entries remain GPT-5.4-based, creating a mismatch between the new default family and the suggested
example overrides. The local installation doc was updated to require GPT-5.5 access, reinforcing
that the default expectation has shifted.

pr_agent/settings/configuration.toml[7-10]
docs/docs/installation/locally.md[1-4]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

### Issue description
The default model was updated to GPT-5.5, but the commented example config lines for `model_reasoning` and `model_weak` still reference GPT-5.4 models, which can confuse users and cause accidental pinning to older models.

### Issue Context
`pr_agent/settings/configuration.toml` is used as the defaults/template config; users often uncomment the example lines rather than choosing models from scratch.

### Fix Focus Areas
- pr_agent/settings/configuration.toml[7-10]

### Suggested change
Update the commented example model names to align with the new default model family (GPT-5.5), or otherwise explicitly note they are examples and may be older than the default.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---