promptdojo_
Checkpoint

One last thing before we move on. Same surface as a write step — but the lesson doesn't complete until this passes.

Final drill. Audit a real-shaped agent stack. Write audit_stack(projects) that takes a list of project profiles and returns a dict with two fields:

  • recommendations: dict mapping each project's name to its recommendation string ("framework-free", "vercel-ai-sdk", "langgraph", or "crewai").
  • framework_free_count: int — how many projects stay framework-free.

Use the SAME rule order as step 7 (first match wins):

  1. team_size <= 2 AND time_to_first_demo_days <= 14 → "framework-free"
  2. needs_streaming AND language in ("typescript", "both") → "vercel-ai-sdk"
  3. needs_durable_state AND language in ("python", "both") → "langgraph"
  4. else → "crewai"

Five projects run. Expected output:

Solo MVP             -> framework-free
Chat widget          -> vercel-ai-sdk
Research agent       -> langgraph
Newsroom sim         -> crewai
Pair scripting tool  -> framework-free
framework-free count: 2

this step needs the editor

on desktop today; in the app (coming soon). save your spot and we'll bring you back here when you're ready.

save my spot follow @TFisPython for the app launch

open this same url on a laptop to keep going today.