feat(server): run managed Ollama on Linux and skip CUDA downloads on install (P1.3b)

- Pinned Ollama 0.32.3 (and rollback 0.32.0) Linux x64/arm64 artifacts with
  their published SHA-256 and sizes; .tar.zst archives are decompressed by
  Node 22's zstd and unpacked by the system tar.
- install.sh sets ONNXRUNTIME_NODE_INSTALL_CUDA=skip: onnxruntime-node
  otherwise downloads CUDA libraries on every Linux x64 install.
- Drop the root sqlite-vec-windows-x64 dependency; the sqlite-vec package
  already carries per-platform binaries.
This commit is contained in:
shumeiko.it
2026-10-24 12:04:18 +02:00
parent f4d2a62b24
commit 55f42a1077
7 changed files with 135 additions and 8 deletions

View File

@@ -1,7 +1,7 @@
{
"name": "@waggle/cli",
"version": "0.1.0",
"description": "Waggle CLI \u2014 interactive AI agent with persistent memory",
"description": "Waggle CLI — interactive AI agent with persistent memory",
"type": "module",
"main": "dist/index.js",
"types": "dist/index.d.ts",