coze-dev / coze-dev/coze-loop

[FEATURE] Add Export to Excel for Prompt Experiment Results

Open
#140 0 comments 0 reactions 0 assignees View on GitHub
enhancement
Dominant language
Go
Stars
5.7k
Forks
795
Avg merge
1d 4h
Merged PRs (30d)
29

Description

### 📋 Please ensure that:

- [x] I have searched existing issues to avoid duplicates
- [x] I have provided a clear problem statement and solution
- [x] I understand this is a feature request and not a bug report
- [x] I am willing to help implement this feature if needed
- [x] I have submitted this feature request in English (otherwise it will not be processed)

### 🎯 Problem Statement

Currently, the results of prompt experiments can only be viewed within the platform. For many scenarios, we need to further process these results outside the system.

It would be very useful to have an “Export to Excel” feature for prompt experiment results. This would allow users to:

Perform manual annotation and review of results.

Define custom statistical metrics (beyond what the platform currently provides).

Calculate custom scores for experiments based on their own evaluation criteria.

Share experiment results easily with other team members who may not have direct access to the platform.

### 💡 Proposed Solution

Add an option to export prompt experiment results in Excel (.xlsx) format.

Include key experiment metadata in the export (e.g., experiment name, prompt, input data, model config, output, system evaluation score, etc.).

Ensure the exported format is structured in a way that facilitates annotation and further analysis (e.g., rows = test cases, columns = attributes).

### 📋 Use Cases

Bulk Evaluation Review
After submitting 500 test cases for evaluation, users need to review all results in one place. An Excel export would present each evaluation case as a single row, with structured columns for analysis.

Suggested Table Structure (Excel headers)

ID | Input Data | Model Parameters | Final Prompt (after concatenation) | Model Output | Evaluator Result

This structure makes it easy to filter, sort, and analyze across hundreds of cases.

### ⚡ Priority

High - Would significantly improve my experience

### 🔧 Component

Evaluation

### 🔄 Alternatives Considered

_No response_

### 🎨 Mockups/Designs

_No response_

### ⚙️ Technical Details

_No response_

### ✅ Acceptance Criteria

_No response_

### 📝 Additional Context

_No response_

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.