Generate a Test Plan from a Feature Spec

Here is a feature spec:

{{feature_spec}}

Write a test plan covering:
1. Happy-path cases — the core behavior working as intended.
2. Boundary/edge cases — empty input, maximum size, zero, negative numbers, unicode,
   concurrent access, whichever apply to this feature specifically.
3. Failure/error cases — what should happen when a dependency is unavailable, input
   is invalid, or a precondition isn't met.
4. Cases that touch existing behavior — anything this feature could plausibly break
   that isn't obviously "this feature" (a shared cache, a rate limit, a permission
   check).

For each case, give: a short name, the setup/input, and the expected outcome. Flag
any case where the spec doesn't actually say what the expected outcome should be —
those are spec gaps, not test-writing problems, and should be called out separately
rather than guessed at.