May 2

1. 18:03

   fix: switch model back to gemini-2.5-flash per user request

   Sanket22g
   a3ec5c7
2. 17:47

   fix: initialize raw variable to avoid UnboundLocalError during 429 quota exceptions, and switch to gemini-2.0-flash for higher daily limit

   Sanket22g
   8087b81
3. 17:41

   fix: use strict=False in json.loads to gracefully handle literal newlines

   Sanket22g
   b625b13
4. 17:37

   chore: add raw output to fallback rationale for debugging

   Sanket22g
   5e9094f
5. 17:34

   fix: prompt formatting to avoid literal newlines in json strings

   Sanket22g
   2f4e961
6. 17:31

   feat: migrate to google-genai SDK for reliable JSON response

   Sanket22g
   6db567d
7. 17:29

   fix: robust JSON extraction for gemini-2.5-flash responses

   Sanket22g
   5bd351c
8. 17:26

   feat: upgrade to gemini-2.5-flash model

   Sanket22g
   2564428
9. 17:06

   chore: restart to pick up new GEMINI_API_KEY secret

   Sanket22g
   b53e7b7
10. 17:00

    fix: pin google-generativeai==0.8.3, system_instruction in model constructor

    Sanket22g
    616229b
11. 16:58

    fix: move system_instruction to model constructor (Gemini SDK compat)

    Sanket22g
    9536fd5
12. 16:48

    feat: Vera bot - FastAPI + Gemini 2.0 Flash, 4-context composer, auto-reply detection

    Sanket22g
    69eb640