update docs

docs/docs/get-started/quick-tour.md

@@ -77,7 +77,7 @@ First of all, we need to get our `trajectory`, or series of actions, correspondi
 To do this, we initialize a WebAgent and give it an objective of testing a natural language `scenario`.
 
 ```python
-from lavague.core.agents import WebAgent
+from lavague import WebAgent
 
 # Create Web Agent
 agent = WebAgent(api_key="")

docs/docs/learn/trajectory.md

@@ -0,0 +1,21 @@
+# Trajectories
+
+Trajectories are classes returned by LaVague agents containing information about an agent run and the list of actions performed.
+
+They contain the following attributes:
+- `start_url`: The URL provided as input for this run
+- `objective`: The text objective that was given as input for this run
+- `run_id`: A unique run id
+- `output`: Text output where expected
+- `actions`: The series of [actions](./actions.md) generated during this run
+- `status`: Whether the agent run was `completed` or `failed`
+
+![trajectory](https://raw.githubusercontent.com/lavague-ai/LaVague/drafting-some-docs/docs/assets/trajectory.png)
+
+## Integrations
+
+This Trajectory can be passed to `exporter` functions to be converted into replayable automation code in the format required your use case.
+
+![exporter](https://raw.githubusercontent.com/lavague-ai/LaVague/drafting-some-docs/docs/assets/trajectory-export.png)
+
+Currently, we only provide a `PyTest exporter`, but we plan to work with the community to build exporters to cover a wide range of use cases.

docs/index.md

@@ -60,11 +60,11 @@ pip install lavague
 You're now ready to start using LaVague Web Agents to perform tasks on the websites of your choice:
 
 ```python
-from lavague.core.agents import WebAgent
+from lavague import WebAgent
 
 url = "https://huggingface.co/"
-objective = "Return a list of the top 5 trending models on HuggingFace"
-agent = WebAgent()
+obj = "Return a list of the top 5 trending models on HuggingFace"
+agent = WebAgent(api_key="")
 ret = agent.run(url=url, objective=obj)
 
 # Print output