feat(assessment): Configuration and Versioning

[vprashrex]

Issue

Closes #1098

Summary

• Before: There was no configuration and versioning for assessment module.
• Now: Existing configuration and versioning module is extended to support assessment configuration and versioning.
• The new implementation required creating the config handling logic.
• Validation checks were introduced to discriminate the default or assessment tag.

Checklist

Before submitting a pull request, please ensure that you mark these task.

• Ran fastapi run --reload app/main.py or docker compose up in the repository root and test.
• If you've fixed a bug or added code that is tested and has test cases.

Notes

This is how the assessment config blob will look like:

{
  "name": "sst-grade10-grader",
  "description": "Rubric-based grader for Class 10 Social Science (SST) answers",
  "tag": "ASSESSMENT",
  "commit_message": "initial",
  "config_blob": {
    "pre_filters": {
      "topic_relevance": {
        "prompt": "Question: {{question}}\nStudent answer: {{student_answer}}\nRate 0-1 how on-topic the answer is for this SST question. Off-topic/blank answers score near 0."
      }
    },
    "assessment": {
      "system_prompt": "You are a CBSE Class 10 Social Science examiner. Grade the STUDENT ANSWER strictly against the MARKING RUBRIC and MODEL ANSWER...\n\nQUESTION ({{max_marks}} marks): {{question}}\nMODEL ANSWER: {{model_answer}}\nSTUDENT ANSWER: {{student_answer}}\n\nRUBRIC:\n- content_accuracy (40%)\n- relevance (25%)\n- terms_and_examples (20%)\n- structure (15%)\nSum into awarded_marks (0..{{max_marks}}).",
      "json_schema": {
        "type": "object",
        "properties": {
          "awarded_marks": { "type": "number" },
          "rubric_breakdown": {
            "type": "object",
            "properties": {
              "content_accuracy":   { "type": "object", "properties": { "score": {"type":"number"}, "comment": {"type":"string"} }, "required": ["score","comment"] },
              "relevance":          { "type": "object", "properties": { "score": {"type":"number"}, "comment": {"type":"string"} }, "required": ["score","comment"] },
              "terms_and_examples": { "type": "object", "properties": { "score": {"type":"number"}, "comment": {"type":"string"} }, "required": ["score","comment"] },
              "structure":          { "type": "object", "properties": { "score": {"type":"number"}, "comment": {"type":"string"} }, "required": ["score","comment"] }
            },
            "required": ["content_accuracy","relevance","terms_and_examples","structure"]
          },
          "keywords_missing": { "type": "array", "items": { "type": "string" } },
          "overall_feedback": { "type": "string" }
        },
        "required": ["awarded_marks", "rubric_breakdown", "overall_feedback"]
      },
      "model": { "provider": "openai", "model": "gpt-4o", "temperature": 0.0 }
    }
  }
}

[coderabbitai]

Important

Review skipped

Auto reviews are limited based on label configuration.

🏷️ Required labels (at least one) (1)

• ready-for-review

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 859b775c-8d68-4c5a-a4f6-9b4c3d4e4421

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}