indifferentketchup
c2c4f78a26
- Add ai@^6 and @ai-sdk/openai-compatible@^2 to apps/server.
- New services/inference/provider.ts: createOpenAICompatible against
llama-swap (baseURL threaded from config.LLAMA_SWAP_URL, cached per
baseURL). No apiKey — Authelia + Tailscale gate llama-swap, not keys.
- streamCompletion rewritten as an adapter over streamText. AI SDK
fullStream parts (text-delta, tool-call, finish, error) map back to
the legacy {content?, tool_calls?, finishReason} StreamResult shape
that executeStreamPhase already consumes. No layer above
streamCompletion changes.
- toModelMessages converts BooCode's OpenAI-shaped history to AI SDK
ModelMessage[]; tool messages need toolName which we look up by
scanning earlier assistant tool_calls for the matching id.
- buildAiTools wraps BooCode's JSON-schema tool defs via
tool({ inputSchema: jsonSchema(parameters) }) with NO execute —
BooCode dispatches tools in tool-phase.ts, not the AI SDK loop.
- XML fallback parser preserved as-is — qwen3.6 still emits XML tool
calls in text content that the structured tool-call layer misses.
- reasoning-delta parts dropped with a debug-level counter — captured
properly in v1.13.1-C.
- Abort path: streamText({ abortSignal }) wires ctx.signal through, but
AI SDK v6 swallows the abort (fullStream iterator exits cleanly
rather than throwing). Post-iteration `if (signal?.aborted) throw` so
handleAbortOrError owns the row and writes status='cancelled'. Caught
by smoke D; would have shipped as status='complete' on stop otherwise.
- Usage frame reads result.usage (inputTokens / outputTokens v6 names)
AFTER stream drain. Single trailing publish through the existing 500ms
throttle. Known regression: ChatThroughput's live mid-stream tick
(v1.12.2) is gone — it now shows a single value at stream end.
TODO(v1.13.1-followup): interpolate outputTokens during streaming
via a delta-cadence counter (e.g. part.text.length/4 token proxy)
and publish every 500ms; reconcile against result.usage at finish.
- Write-path dual-write from v1.13.0 unaffected.
Read path stays on JSON columns. v1.13.1-B flips reads to message_parts.
Smoke verified end-to-end against running container:
- A. Plain text: status='complete', 1 text part.
- B. Single tool prompt → multi-tool chain (4 calls): every assistant
with tool_calls has 2 parts (text+tool_call), every tool row has
1 part (tool_result).
- C. Multi-step covered by B's chain.
- D. Stop mid-stream: status='cancelled' written via handleAbortOrError
after the post-iteration abort throw.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
boocode
Self-hosted single-user developer chat app. v1: chat only.
Stack
- Node 20, Fastify, postgres (porsager/postgres), ws, zod
- React 18, Vite, TypeScript, Tailwind v4, shadcn/ui
- Postgres 16
- pnpm workspaces
Layout
apps/server— Fastify API + WebSocket + inference loop + file-read toolsapps/web— React frontend; served by Fastify in production, Vite in dev
Local dev
Requires Node 20, pnpm, Docker (for Postgres), and ripgrep.
# install
pnpm install
# bring up postgres only
cp .env.example .env
# edit POSTGRES_PASSWORD if you like; default DATABASE_URL points at the container
docker compose up -d boocode_db
# run server (port 3000) and web (port 5173) in two shells
DATABASE_URL=postgres://boocode:devpass@127.0.0.1:5500/boocode \
LLAMA_SWAP_URL=http://100.101.41.16:8401 \
pnpm dev:server
pnpm dev:web
The Vite dev server proxies /api and /api/ws/* to the Fastify backend with a
synthetic Remote-User: sam header so the Authelia auth layer can be skipped
during development.
Production
cd /opt/boocode
docker compose up --build -d
Binds to 100.114.205.53:9500 (Tailscale). Authelia is expected to gate the
upstream and inject Remote-User. Postgres binds loopback only.
What v1 has
Project sidebar, sessions per project, chat with streaming responses over
WebSocket, four file-read tools scoped to the project root (view_file,
list_dir, grep, find_files), and a model picker driven by llama-swap's
/v1/models.
What v1 does not have lives in v2 (terminal pane) and v3 (Coder pane).