Skip to content

Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. πŸ”₯

docs(readme): add Interact via Prompting subsection, replace em dashes

firecrawl-spring[bot]1a05039a761a
diff --git a/README.md b/README.md
index f38a70474..9664f3237 100644
--- a/README.md
+++ b/README.md
@@ -295,7 +295,11 @@ results = firecrawl.search(
## Interact
-Scrape a page, then interact with it β€” click buttons, fill forms, extract dynamic content, or navigate deeper. Use natural language prompts or run code for full control. [Learn more β†’](https://docs.firecrawl.dev/features/interact)
+Scrape a page, then interact with it - click buttons, fill forms, extract dynamic content, or navigate deeper. Use natural language prompts or run code for full control. [Learn more β†’](https://docs.firecrawl.dev/features/interact)
+
+### Interact via Prompting
+
+Describe what you want in natural language. The agent will click, type, scroll, and extract data automatically.
```python
from firecrawl import Firecrawl
@@ -317,7 +321,7 @@ app.stop_interaction(scrape_id)
### Run Code in the Browser
-For full control, execute Playwright code directly β€” `page` is already connected:
+For full control, execute Playwright code directly - `page` is already connected:
```javascript
import Firecrawl from '@mendable/firecrawl-js';
@@ -357,7 +361,7 @@ scrape_id = result.metadata["scrapeId"]
app.interact(scrape_id, prompt="Fill in [email protected] and password, then click Login")
app.stop_interaction(scrape_id)
-# Next session β€” already logged in
+# Next session - already logged in
result = app.scrape(
"https://app.example.com/dashboard",
formats=["markdown"],