Add proxy configuration tips, wandb integration and refine task instructions

[Flemington8]

Introduce Weights & Biases integration for session logging and orchestrator initialization. Update README and llm.py to support .env file for API key management. Refine task descriptions and instructions for smaller models to enhance format consistency and reduce errors. Include proxy configuration tips for AIOpsLab users.

[Copilot]

Pull Request Overview

This PR adds a documentation section with proxy configuration tips for AIOpsLab users to help avoid issues when running VLLM and the LLM agent locally.

• Introduces a "[Tips]" section in README.md
• Provides recommendations on avoiding the use of the 172.17.0.1 address and handling SOCKS5 proxy connections
• Advises on setting an environment variable (no_proxy) to prevent proxying localhost

[Copilot]

Pull Request Overview

This PR refines task instructions across multiple task modules and adds proxy configuration tips to improve the developer experience when using AIOpsLab. Key changes include updated API call formatting instructions, language adjustments for task descriptions, and additional documentation on proxy settings in the README.

Reviewed Changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
clients/utils/templates.py	Added important instructions for API call formatting
aiopslab/orchestrator/tasks/mitigation.py	Refined task description and API call examples
aiopslab/orchestrator/tasks/localization.py	Updated task instructions and clarified expected submission format
aiopslab/orchestrator/tasks/detection.py	Minor text updates to align instruction language
aiopslab/orchestrator/tasks/analysis.py	Revised task instructions and submission guidelines
aiopslab/orchestrator/actions/detection.py	Updated parameter descriptions in the docstring
README.md	Added proxy configuration tips and guidance for SOCKS5 usage

Comments suppressed due to low confidence (1)

aiopslab/orchestrator/tasks/localization.py:37

• Consider clarifying the expected type by using list[str] (if appropriate) instead of list[] for a more explicit type description.

2. list[]: an empty list if no faults were detected

[Flemington8]

=== Qwen2.5-Coder-3B-Instruct Error Parsing Analysis (Before Modification) ===
Total JSON files: 54
Total trace steps: 536
Total 'Error parsing' responses: 460
Error parsing ratio: 85.82%

=== Qwen2.5-Coder-3B-Instruct Error Parsing Analysis (After Modification) ===
Total JSON files: 207
Total trace steps: 924
Total 'Error parsing' responses: 1
Error parsing ratio: 0.11%

As you can see, the refinement of task descriptions and instructions could decrease the Error parsing ratio significantly.

[tianyin]

@HacksonClark Can you review this PR?

Thanks for the contribution @Flemington8 !

[HacksonClark]

@Flemington8 Left some comments for you, overall looks great! I just think it's better if we make use_wandb into a .env variable instead of an argument, it keeps the code a lot cleaner. Once that is done I think this will be ready to merge.

[Flemington8]

@Flemington8 Left some comments for you, overall looks great! I just think it's better if we make use_wandb into a .env variable instead of an argument, it keeps the code a lot cleaner. Once that is done I think this will be ready to merge.

Great, thank you for your suggestion. This could make the code logic more concise!

[Flemington8]

I have updated my PR follow your suggestion, is there any further modification I need to apply? @HacksonClark

[HacksonClark]

Great job!