fix: updating a few metrics to take into account computation at the agent level

[Castorche]

Description

This PR updates a few of the metrics to compute them at the agent level.

Type of Change

• Bugfix
• New Feature
• Breaking Change
• Refactor
• Documentation
• Other (please describe)

Checklist

• I have read the contributing guidelines
• Existing issues have been referenced (where applicable)
• I have verified this change is not present in other open pull requests
• Functionality is documented
• All code style checks pass
• New code contribution is covered by automated tests
• All new and existing tests pass

[k-payumo]

The processor.py should be updated imo to iterate through the relevant agent_names, then apply these metrics to spans belonging to that respective agent. Similar to the changes that are proposed in the deepeval adapter, but should be abstracted higher level to how the MCE iterates through entities.

We may need to spend to time to determine which metrics we should consider as a complete E2E evaluation of as MAS, vs independent evaluations of individual agents.

[k-payumo]

Minor comment on redacting of images. Doesn't need to be removed here but just a heads up, will have a more holistic filter that removes unused metadata fields such as

PAYLOAD_FILTER_KEYS: Set[str] = {
"embedding",
"image_url",
"url",
"class",
"module",
"config_specs",
"lc_attributes",
"lc_secrets",
"model_computed_fields",
"model_config",
"optional_variables",
"input_types",
"cache",
"verbose",
"callbacks",
"tags",
}