Graph embeddings for checklists

[AlexMikhalev]

New post: https://terraphim.ai/posts/graph-embeddings-checklists/

A checklist is an enumerated set of terms that must be checked. Vector embeddings return fuzzy neighbours; a knowledge graph can enumerate and name the edge that fired. The post walks the checklist lifecycle around an LLM call: terraphim-grep with a role thesaurus expands the query into concept terms before it (eight concepts for 'quality gate' in 53 ms), terraphim-agent hook returns a findings checklist after it, and learn/memory feeds corrections back into the synonyms.

All quoted outputs are real, including one honest caveat about terraphim-agent replace in this build. The site repo now carries a .terraphim/ directory (kg concepts, rewrite checklist, 39-entry thesaurus) so the commands reproduce.

Which checklist in your agent setup would you enforce with an automaton rather than a prompt?