dogfood-run

name	dogfood-run
description	Run dogfooding iterations to test the Sharpy compiler
argument-hint	[number_of_iterations]

Run the Sharpy dogfooding pipeline to generate, compile, and verify AI-generated test cases.

Steps

1. Parse iteration count from $ARGUMENTS. If empty or non-numeric, default to 5.

2. Run the dogfooding tool in the background using Bash with run_in_background: true:

   cd build_tools && python3 -m sharpy_dogfood run -n <N> --verbose 2>&1
   

   Replace <N> with the parsed iteration count.

3. Inform the user that dogfooding is running in the background. Each iteration involves AI code generation + compilation + execution, so estimate roughly 1-2 minutes per iteration.

4. Wait for completion by reading the background task output.

5. Present results. After the task completes:

   • Read dogfood_output/SUMMARY.md (relative to the project root /Users/anton/Documents/github/sharpy/) and present its contents to the user.
   • If the summary mentions failures or issues, suggest running /dogfood-analyze to classify them by root cause.