fix output dir logging

[mikasenghaas]

Description

two small fixes to output_dir in vf-eval:

• logging dir is also relative to output_dir
• can be set via cli (not only toml)

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Test improvement

Testing

• All existing tests pass when running uv run pytest locally.
• New tests have been added to cover the changes

Checklist

• My code follows the style guidelines of this project as outlined in AGENTS.md
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published

Additional Notes

---

Note

Medium Risk
Moderate risk because it changes evaluation output/log path wiring and env-server logging initialization, which could affect where logs are written or whether logging is configured as expected across debug/non-debug runs.

Overview
Adds --output-dir (-o) to prime eval run so CLI-only evaluations can override the base directory used for saved results and logs, and documents the new flag/behavior.

Fixes env-server logging to consistently write eval.log under the computed results_path (including in debug mode) and always register the log file callback, and refactors EnvServer logging setup to only configure file logging when log_file is explicitly provided (removing the implicit logs/{env_id}.log default).

^{Written by Cursor Bugbot for commit bc038ee. This will update automatically on new commits. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}