MCP Content Source Config

[mrchantey]

We need to be able to configure and share the source info for each crate for rag indexing. Now that the KnownSources can be serialized and passed in we can work on a config format.

Something like this might work with the current ContentSource type.

# beet.toml


[content-sources]

[[include]]
path = "https://github.com/mrchantey/beet/crate-sources/bevy-and-friends.toml"

[[include]]
path = "./my-other-sources.toml"

[[source]]
crate-name = "bevy"
crate-version = "0.16.0"
content-type = "guides"
git-url = "https://github.com/bevyengine/bevy-website"
git-commit = "abc123"
filter = "*/*.md"

[[source]]
crate-name = "bevy"
crate-version = "0.16.0"
content-type = "examples"
git-url = "https://github.com/bevyengine/bevy"
git-commit = "abc123"
git-branch = "dev"
filter = "**/examples/**/*.rs"
split-text.chunk-range.min=1500
split-text.chunk-range.max=3000

We should probably rename KnownSources to something like ContentSourceConfig.