Search endpoint returns one fewer result than expected (off-by-one)

[anxkhn]

Context

The GET /search endpoint has an off-by-one error in its pagination slicing. When the number of matching results equals or exceeds the limit, it returns one fewer result than it should.

Steps to reproduce

1. Create 5 profiles with usernames starting with "a" (alice, alex, anna, amy, adam).
2. Search with GET /search?q=a&limit=3.
3. The total field correctly shows 5, but results contains only 2 items instead of 3.

You can also reproduce with a test:

def test_search_pagination():
    # (add 5 profiles with "a" in username)
    response = client.get("/search?q=a&limit=3")
    assert len(response.json()["results"]) == 3  # FAILS: returns 2

Expected behavior

With limit=3, the endpoint should return up to 3 matching results.

Actual behavior

With limit=3, the endpoint returns only 2 results because the slice uses offset + limit - 1 instead of offset + limit.

Files

• app/main.py -- the search_profiles function, the return statement with the slice

Acceptance criteria

• GET /search?q=...&limit=N returns up to N results (not N-1)
• Existing search tests still pass
• The total field remains unchanged (full count, not sliced count)

Suggested approach

1. Open app/main.py and find the search_profiles function.
2. Find the line: results[offset : offset + limit - 1]
3. Change it to: results[offset : offset + limit]
4. Run pytest tests/test_search.py -v.