Test Log - _07_image_extraction

Tested 2026-05-19 against gemini-3.5-flash (Gemini), agno 2.6.8.

basic.py

Status: PASS

Description: Extract a Scene (subject, setting, time of day, colors, objects) from a photo of Krakow's St. Mary's Basilica.

Result: All fields populated correctly.

Note: Test image switched from a Wikimedia URL (Gemini can't fetch those) to agno-public.s3.amazonaws.com/images/krakow_mariacki.jpg.

---

ocr_fields.py

Status: PASS

Description: OCR a text-heavy image into typed fields (primary text, secondary text, color scheme).

Result: Reads the Agno wordmark and body copy from the intro image into the typed schema.

Note: Input switched from a Wikimedia "Welcome to Las Vegas" sign (Gemini can't fetch those URLs) to agno-public.s3.us-east-1.amazonaws.com/images/agno-intro.png. The schema is still about extracting text from an image — content changed from a sign to a marketing graphic.

---

with_confidence.py

Status: PASS

Description: Same task with ConfidentStr / ConfidentList wrapping each output (fjord landscape from the gstatic gallery).

Result: All fields populated with high confidence; nested confidence wrappers serialize correctly.

---