How do you use Synto? Share your use cases, successes & failures

[kytmanov]

Hey everyone,

I'd love to hear how people are actually using Synto in real life.

Whether it's extracting concepts from research papers, building personal knowledge bases, processing documentation, or anything else — please share:

• What you're using it for
• Your typical workflow / source material
• What’s working well
• Any frustrations or failures you've run into

No pressure to only share wins — honest feedback is very welcome too. This will help the project and the community a lot.

Looking forward to your stories! 🚀

[bstovej]

I currently use it as an extension to my Obsidian vault for specific knowledge domain. I would prefer that it can integrates with my Obsidian vault, which uses a different (more traditional) folder structure (00_Inbox, 01_Topics, ... 03_Resources, etc.), which rightfully, 00_Inbox should map to raw and 01_Topics or a MOC folder should map to wiki/concept, etc., but synto currently does not allow for such mapping. Would be great if such flexibility can be provided by defining the mapping in the synto.toml file. Another ask I have is the flexibility to merge two or more concepts (e.g., EU AI ACT, EU_AI_ACT, EU-AI-ACT, which could be picked up as three different concepts but they are really the same thing.) Also, can I delete any articles raw/sources/concepts from the wiki and synto will take care to re-sync the repo on the next run?

[kytmanov]

@bstovej Thanks for sharing, these are all good ones.

Merging duplicate concepts are already covered. Since v0.6.0 every concept has a stable identity, and ingest normalizes case + separators, so EU AI ACT / EU_AI_ACT / EU-AI-ACT usually collapse into one concept on their own. If they didn't (happens on older vaults), you can merge them by hand:

• synto concept inspect "EU AI ACT" to see suggestions
• synto concept merge EU_AI_ACT "EU AI ACT" to fold one into the other (links get rewritten, the extra article is retired)
• synto concept unmerge if you want to undo it

Folder mapping in synto.toml - not yet, but I get it. Right now the layout (raw/, wiki/) is fixed, and it's baked in deeper than a config alias - synto stores those paths in its database and figures out article types from the folder - so doing it properly means a path-config layer + a migration. Putting it on the list. For now the clean option is to run synto on a dedicated vault/subfolder instead of overlaying it on 00_Inbox / 01_Topics directly.

Deleting files and re-syncing - not there yet but it should be. Today if you delete a note from raw/ or an article from wiki/, synto leaves the leftover bookkeeping behind and won't clean up the derived concepts/articles. lint/doctor flag orphans and maintain --fix repairs links, but nothing prunes deleted sources yet.

Logged 1 and 3. Appreciate the detailed notes!

[bstovej]

Hi @kytmanov, thanks for you reply and advise. I decided last night to use symlink to map raw and wiki to my vault folders and so far it seems to work fine. I decided to link my resource folder (03_Resources) to raw instead of 00_Inbox as I noted that files dropped into raw don't get moved to another folder, which conflicts with the idea of an inbox. The inbox (in my context) is a holding area that I would clear from time to time, whereas my resource folder is more like a long term storage area. Cheers!