Support regenerate for execute API

[Hailong-am]

Description

Support regenerate on existing interaction, regenerate_interaction_id and memory_id must be provide for regenerate case.

POST http://localhost:9200/_plugins/_ml/agents/MBWOSIwBAnQBpyKtmAkn/_execute
{
  "parameters": {
    "memory_id": "A5CgV4wBiufF8-BkxC0r",
    "regenerate_interaction_id": "BJCgV4wBiufF8-BkxC2b",
    "verbose": false
  }
}

Response

{
    "inference_results": [
        {
            "output": [
                {
                    "name": "response",
                    "dataAsMap": {
                        "response": "Your OpenSearch cluster contains many security audit log indices from late 2023, with most replicas in a yellow status. The cluster and index storage sizes and document counts appear normal.",
                        "additional_info": {}
                    }
                },
                {
                    "name": "QuestionSuggestor",
                    "result": "[\"What are some ways I can analyze these security audit logs further?\",\"Can you help me understand why the replicas are in yellow status?\"]"
                }
            ]
        }
    ]
}

Issues Resolved

[List any issues this PR will resolve]

Check List

• New functionality includes testing.
  • All tests pass
• New functionality has been documented.
  • New functionality has javadoc added
• Commits are signed per the DCO using --signoff

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.

[codecov]

Codecov Report

Attention: 57 lines in your changes are missing coverage. Please review.

Comparison is base (9155655) 69.78% compared to head (5c19d90) 69.56%.

Files	Patch %	Lines
...g/opensearch/ml/engine/memory/MLMemoryManager.java	0.00%	35 Missing ⚠️
...ch/ml/engine/algorithms/agent/MLAgentExecutor.java	53.84%	16 Missing and 2 partials ⚠️
.../ml/engine/algorithms/agent/MLChatAgentRunner.java	69.23%	4 Missing ⚠️

Additional details and impacted files

@@                        Coverage Diff                        @@
##             feature/agent_framework_dev    #1709      +/-   ##
=================================================================
- Coverage                          69.78%   69.56%   -0.23%     
- Complexity                          2660     2667       +7     
=================================================================
  Files                                244      244              
  Lines                              12977    13036      +59     
  Branches                            1305     1309       +4     
=================================================================
+ Hits                                9056     9068      +12     
- Misses                              3303     3347      +44     
- Partials                             618      621       +3     

Flag	Coverage Δ
ml-commons	69.56% <34.48%> (-0.23%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[arjunkumargiri]

Can interaction regeneration logic be handled in MLAgenExecutor instead of Chat agent? In future there could be new agent types and we should not re-implement regeneration logic in each agent type

[Hailong-am]

@arjunkumargiri Instead of put logic in MLAgenExecutor, we can have a base class for all kinds of agent runner, put common logic in that class would make more sense.

[arjunkumargiri]

Agreed, makes sense.

[Hailong-am]

@arjunkumargiri Since the root interaction been moved to MLAgentRunner,regenerate is coupled with root interaction, move the logic to that class also.

[dhrubo-os]

What if both are null? Do we still want to continue?

[Hailong-am]

if regenerateInteractionId is null, it's a normal flow, memory will automatically create if needed

The base branch was changed.

[Hailong-am]

done in #1816