Exercises

❯ cat documents/

• -rw-r--r-- Build an agent from scratch Exercise 2026-08-08

  An agent from first principles — raw LLM API, schema-validated tools, the loop, tracing and evals — no framework, until the magic is demonstrably a while-loop.

• -rw-r--r-- Build an agent from scratch — subject Subject 2026-08-08

  The spec for the no-framework agent — raw HTTP to the LLM API, a self-defined tool-call schema, three-plus gated tools, a bounded loop, full tracing, and a scored eval set.

• -rw-r--r-- RAG over the knowledge base Exercise 2026-08-08

  A RAG pipeline over this very repo — structure-aware chunking, embeddings, hybrid retrieval, cited generation — with evals that prove where the bottleneck is.

• -rw-r--r-- RAG over the knowledge base — subject Subject 2026-08-08

  The work statement for the RAG build over tech-kb — structure-aware chunking rules, hybrid retrieval with rank fusion, path-citing generation, and a hit-rate-scored retrieval eval set.