Typedef workshop: building a context graph layer for data and coding agents
AI Engineer · youtube · 2026-10-11
Typedef's Yoni Michael and Brandon Callender demonstrate a context layer for agents: text search loses identity, context graphs preserve it. Valid SQL can still yield wrong answers, so the data side computes grain, lineage, and relationships from transformation code and validates join assumptions. On the code side, Tree-sitter, Rustdoc, and Neo4j build a queryable symbol graph (built over Qdrant crates) so agents can ask blast-radius questions instead of grepping. Key lesson: agents may still choose grep—tool naming, output formats, and prompting are part of the interface, and evals must measure tool usage. Closes with incremental indexing, Python's harder symbol resolution, and Fenic semantic operations. Full timestamps and repo included.
More from coding & agent
- O'Reilly 'Agent Memory' book enters early release with first two chapters live — danielrock · 2026-10-11
- saccade: A Local-First Python Library for Video Transcripts, Frames, and Semantic Search — Dapper_Ad599 · 2026-10-11
- Indie dev migrates agent tool Toolfolio to Whop payments, praising its dead-simple CLI — eptwts · 2026-10-11
- Developer flags Claude web interface restriction on uploading new skills, multi-file skills blocked — strickvl · 2026-10-11
- Users want X Money integration to give AI agents dedicated, capped budget envelopes — RachelVT42 · 2026-10-11
- LangChain's CEO on the open question of identity for internal company agents — hwchase17 · 2026-10-11