README
This category is where the Chat Import Pipeline gets designed
and understood.
The Chat Import Pipeline is the mechanism by which AI
conversations — currently Claude, extensible to others — are
brought into LOGOS as first-class content. Not archived as
flat files. Not summarized. Imported with full structure,
metadata, and provenance so they are referenceable,
searchable, and feedable to CORTEX.
Why This Matters
A significant portion of the reasoning that shapes LOGOS
happens in AI conversations. Design decisions get made there.
Requirements get refined there. Limitations get discovered
there. If those conversations live only in Claude's interface
they cannot be cross-referenced from a requirement topic,
cannot be searched alongside forum content, and cannot
contribute to CORTEX's growing knowledge of the system.
The pipeline makes AI reasoning a permanent, referenceable
part of the LOGOS knowledge base.
What Lives Here
Constraints and problems — limitations of external
systems discovered during pipeline design. Claude export
gaps, format limitations, metadata that is missing or
incomplete. These are Problem topics — tracked with status,
linked to the LOGOS requirements they generate.
Design discussions — decisions about how the pipeline
works, what it preserves, how it handles edge cases, what
the import interface looks like.
Status tracking — where the pipeline is in development,
what works, what doesn't yet.
Relationship to Other Categories
- LOGOS → Requirements — constraints discovered here
generate requirements there. Each problem topic links to
the requirement it produced. - LOGOS → Talkyard Limitations — parallel structure.
Both are records of constraints in external systems.
Talkyard Limitations records forum platform constraints.
This category records import source constraints. - CORTEX — the pipeline feeds CORTEX. Pipeline quality
directly bounds CORTEX knowledge quality.
Current Constraints
- Progress