Add optional docs_info str to task models

[jluethi]

We should add a docs_info str to the task models. The purpose is that developers can specify a single, markdown formatted string that describes the functionality of their task. This is useful to show under the info button on the task page (public & in Fractal servers). If it's filled out, the docs_info entry in the manifest is filled with that.

Currently, we just show the task docstring as docs_info. We could just improve docstrings to become more user-friendly. But sometimes developers want dev-related info in the docstrings.
Plus, having a single docs_info field in the task models that's filled in the task list will allow us to provide a single summary for compound tasks, instead of combining the docstrings.

Given how these info panels are currently displayed, the content of docs_info can be multiple sentences without an issue.

If we can also adapt to not show the titles quite so large or not default to setting titles based on docstring info parsed at all, that would further help clean up the interface

[tcompa]

I think it will be fairly simple to add this to the manifest-building script. In principle I think that no other changes in fractal-server or fractal-web are needed.

The main question is where this string is stored.

1. We can start by storing the full markdown string in the task_list.py module, but it's not really usable if one wants to build (and preview locally) some more complex markdown file.
2. We can define some custom pattern for the task_list.py objects. E.g. if one sets docs_info="file:relative/path/to/some/file.md, then the manifest-building script use the contents of that file as a docs_info.

Could you provide a basic example for one of the fractal-tasks-core tasks, so that?

[tcompa]

If we can also adapt to not show the titles quite so large or not default to setting titles based on docstring info parsed at all

If the user provides a full string, then that one will fully override the original one.
As per the current automatically-generated ones, I'll change the style of the generated markdown.

[tcompa]

The updates in #876 are now released as fractal-tasks-core 1.3.4, and live at https://fractal-analytics-platform.github.io/fractal_tasks - see e.g.

[jluethi]

Awesome, thanks a lot @tcompa ! This looks great to me, I'll slowly work on getting other tasks packages to adopt this.

Out of curiosity: I someone were to embed an image in the markdown same as we can embed them in READMEs => would that load?

[tcompa]

Out of curiosity: I someone were to embed an image in the markdown same as we can embed them in READMEs => would that load?

Yes, it would. For instance with this task-description file

### Purpose

![Fractal_overview](https://github.com/user-attachments/assets/666c8797-2594-4b8e-b1d2-b43fca66d1df)

- **Segments images using Cellpose models**.
- Supports both **built-in Cellpose models** (shipped with Cellpose) and **user-trained models**.
- Accepts dual image input for segmentation.
- Can process **arbitrary regions of interest (ROIs)**, including whole images, fields of view (FOVs), or masked outputs from prior segmentations, based on corresponding ROI tables.
- Provides access to all advanced Cellpose parameters.
- Allows custom rescaling options per channel, particularly useful for sparse images.


<p align="center">
  <img src="https://raw.githubusercontent.com/fractal-analytics-platform/fractal-logos/refs/heads/main/projects/fractal_server.png" alt="Fractal server" width="400">
</p>

### Limitations
- Compatible only with Cellpose 2.x models; does not yet support 3.x models.

you obtain this info modal

[jluethi]

Awesome, thanks for testing this! :)