chore: update readme

.github/workflows/test.yml

@@ -1,13 +1,12 @@
 name: Test
 
 on:
-  push:
-    branches: [ main ]
   pull_request:
     branches: [ main ]
 
 jobs:
   test:
+    if: !startsWith(github.event.head_commit.message, 'chore')
     runs-on: ubuntu-latest
 
     steps:

README.md

@@ -47,23 +47,23 @@ npm run dev $QUERY
 ```
 
 ## Demo
-> using `gemini-1.5-flash` as llm, `gemini-2.0-flash` leads to much better!
+> was recorded with `gemini-1.5-flash`, the latest `gemini-2.0-flash` leads to much better results!
 
 Query: `"what is the latest blog post's title from jina ai?"`
 3 steps; answer is correct!
-![demo1](demo.gif)
+![demo1](.github/visuals/demo.gif)
 
 Query: `"what is the context length of readerlm-v2?"`
 2 steps; answer is correct!
-![demo1](demo3.gif)
+![demo1](.github/visuals/demo3.gif)
 
 Query: `"list all employees from jina ai that u can find, as many as possible"` 
 11 steps; partially correct! but im not in the list :(
-![demo1](demo2.gif)
+![demo1](.github/visuals/demo2.gif)
 
 Query: `"who will be the biggest competitor of Jina AI"` 
 42 steps; future prediction kind, so it's arguably correct! atm Im not seeing `weaviate` as a competitor, but im open for the future "i told you so" moment.
-![demo1](demo4.gif)
+![demo1](.github/visuals/demo4.gif)
 
 More examples:
 

TODO.md

@@ -1,3 +0,0 @@
-~disable answer after bad attempts for 1 step~
-~into the beast mode when all attempts failed or token budget is exhausted, disable all actions but answer~
-XMLize the prompt