+
[](https://github.com/OpenAdaptAI/OpenAdapt/actions/workflows/main.yml)
[](https://pypi.org/project/openadapt/)
[](https://pypi.org/project/openadapt/)
@@ -57,15 +63,19 @@ flowchart LR
R -->|UI drifted| L{"Resolution ladder re-resolves the target?"}
L -->|resolved| V
L -->|no confident match| H["Halt with a report"]
+ V --> P["Illustrated run report"]
+ H --> P
classDef halt stroke:#b21f2d,stroke-width:2px;
+ classDef ok stroke:#3e6b4f,stroke-width:2px;
class H halt;
+ class V ok;
```
*Text summary (PyPI does not render Mermaid): demonstrate once, compile to a
local program, then replay it deterministically with no model calls. If the UI
matches, the write is verified; if it drifted, the resolution ladder re-resolves
the target, and if it cannot establish a confident match the run halts with a
-report instead of guessing.*
+report instead of guessing. Either way, every run writes an illustrated report.*
---
diff --git a/media/openadapt-hero.svg b/media/openadapt-hero.svg
new file mode 100644
index 000000000..94964ea4c
--- /dev/null
+++ b/media/openadapt-hero.svg
@@ -0,0 +1,117 @@
+
\ No newline at end of file