fix(agent): retry transient LLM HTTP errors

[afjcjsbx]

📝 Description

Fixes LLM request retry handling for transient provider HTTP errors. Previously, OpenRouter/OpenAI-compatible Status: 500 responses could fail the agent turn immediately when no model fallback candidate was available. This PR reuses the existing provider error classifier in the main LLM retry loop so transient HTTP/server errors are retried according to the configured retry policy.

🗣️ Type of Change

• 🐞 Bug fix (non-breaking change which fixes an issue)
• ✨ New feature (non-breaking change which adds functionality)
• 📖 Documentation update
• ⚡ Code refactoring (no functional changes, no api changes)

🤖 AI Code Generation

• 🤖 Fully AI-generated (100% AI, 0% Human)
• 🛠️ Mostly AI-generated (AI draft, Human verified/modified)
• 👨‍💻 Mostly Human-written (Human lead, AI assisted or none)

🔗 Related Issue

Fixes #629

📚 Technical Context (Skip for Docs)

• Reference URL: [BUG] Didn't retry if meet LLM call failed #629
• Reasoning: The provider error classifier already maps transient HTTP statuses such as 500, 502, and 503 to retryable failover reasons. The main LLM retry loop now uses that classifier too, so single-provider runs can retry transient provider failures instead of dropping the task.

🧪 Test Environment

• Hardware:
• OS:
• Model/Provider:
• Channels:

📸 Evidence (Optional)

Click to view Logs/Screenshots

☑️ Checklist

• My code/docs follow the style of this project.
• I have performed a self-review of my own changes.
• I have updated the documentation accordingly.