Run Baseline
Spawn a data-analyst agent for each query the user specifies.
code
Task tool parameters:
- subagent_type: "data-analyst"
- model: "haiku"
- prompt: |
Answer this question about the P&L dataset.
**Query ID:** [e.g., q01, q17, q23 - from tests/test_queries.md]
**Question:** [query text]
**Data file:** data/FUN_company_pl_actuals_dataset.csv
Show your work. If you encounter errors or ambiguity, document them honestly.
When spawning agents, always include the Query ID from the test file (q01-q25).