SmallCode Files

[1.4.0] - 2026-05-29

feat: RAG coding harness with Python scraper + hybrid search (#64)

Adds an opt-in retrieval harness that scrapes snippet-sized code chunks from curated repositories and serves them via hybrid (keyword + vector) search. New surface: bin/rag-index.js + smallcode-rag-index bin and the rag:index npm script (builds .smallcode/rag/index.json from chunks scraped by scripts/rag_scraper.py), src/rag/ (index_store.js, retriever.js, curated_repos.json), docs in docs/rag-harness.md, and test/rag.test.js. No new npm dependencies — the scraper runs via Python.

fix: strict chat templates reject mid-conversation system messages (#62)

Qwen3 / Qwen3.5 chat templates (and other strict templates) under llama.cpp --jinja raise System message must be at the beginning. and llama.cpp returns HTTP 400 — but only when tools are present, since that's when it compiles the template to build a tool-call grammar. SmallCode injects system-role content mid-conversation (clarifier, plan request, planner injection, path-validation warnings, skill activation, compaction summaries), producing a messages array with system entries at positions other than 0.

• New src/session/message_normalizer.js#consolidateSystemMessages() collapses all system-role messages into a single leading system message (preserving order, de-duplicating identical blocks) and emits only non-system turns after it.
• Applied in both request builders (bin/smallcode.js and bin/model_client.js chatCompletion) right before the body is sent, so it catches stray system messages regardless of which path injected them. Verified end-to-end against a Qwen3 model: every tool-bearing request now carries exactly one system message at index 0.
• Test coverage: test/message_normalizer.test.js (9 cases).