championble3 opened a new issue, #45237:
URL: https://github.com/apache/superset/issues/45237

   ### Bug description
   
   Problem: Charts created or edited through the REST API/MCP have no 
query_context, or an outdated one. get_chart_data then rebuilds a generic query 
from form_data. For chart types where the frontend buildQuery adds logic (time 
grain, trendline, period comparison, SQL columns), this returns different 
numbers than the dashboard.
    
   Example: big_number chart with time_grain_sqla: P3M, compare_lag: 1, 
show_trend_line: true. The dashboard shows the QoQ change; get_chart_data 
returns the plain sum of the metric.
    
   Why it is hard to notice: the fallback only calls ctx.warning(...), which 
does not reach the client’s tool result, so an LLM client presents the number 
as correct. Charts with a stale saved context are not flagged at all.
    
   Suggestions: (1) include a warning in the tool result (e.g. in insights) 
when the fallback is used; (2) mention in the tool description that exact 
results need a chart saved from Explore; (3) consider flagging a saved 
query_context older than the chart’s changed_on.
    
   Related: 
[apache/superset#39413](https://github.com/apache/superset/pull/39413) (fixed 
update_chart), 
[apache/superset#45210](https://github.com/apache/superset/issues/45210), 
[#28446](https://github.com/apache/superset/issues/28446).
   
   ### Screenshots/recordings
   
   _No response_
   
   ### Superset version
   
   6.1.0
   
   ### Python version
   
   3.11
   
   ### Node version
   
   16
   
   ### Browser
   
   Chrome
   
   ### Additional context
   
   Observed on 6.1.0 (built-in MCP service) with charts created through the 
API. On current master the fallback still only calls ctx.warning, and insights 
has no entry for it.
   
   ### Checklist
   
   - [x] I have searched Superset docs and Slack and didn't find a solution to 
my problem.
   - [x] I have searched the GitHub issue tracker and didn't find a similar bug 
report.
   - [x] I have checked Superset's logs for errors and if I found a relevant 
Python stacktrace, I included it here as text in the "additional context" 
section.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to