agentscope-ai / agentscope-ai/QwenPaw-Data

[Bug]: Too mush token cost for [Demo PG - GAAP use case ]

Open
#23 2 comments 0 reactions 1 assignee Claimed by @cyruszhang View on GitHub
bug
Dominant language
Python
Stars
74
Forks
11
Avg merge
14h 4m
Merged PRs (30d)
27

Description

### Summary

5,014,326 token cost (about $2) in https://bailian.console.aliyun.com for task
```
datapaw run --datasource-id postgresql-demo-gaap "分析 2026 年 3 月 product X 有效用户的平均 GAAP 值,展示时间趋势,并结合相关 KG 事件解释异常峰值"
```

### Minimal reproduction

minimal demo as README

### Environment

minimal demo as README

### Sanitized logs

```shell

```

### Checklist

- [x] I searched existing issues and discussions.
- [x] I removed secrets, private URLs, and business data.
- [x] This is not a security vulnerability.

Contributor guide

Open the contributing guide

Research direction

Start with the README's minimal demo and the `datapaw run` entry point for `postgresql-demo-gaap`. Trace how the sample query and knowledge-graph explanation consume tokens, then compare the reported usage with the Bailian console for the reproduction command. Done means the demo no longer incurs the reported excessive token cost and the result remains correct.

Written by the indexing model from the issue text.

Assessment

Tech stack
postgresql, python
Domain
ai, data, databases, performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.