Environment

• OS: macOS Apple Silicon (arm64)
• Node.js: v22.x
• Clawdbot: latest

Problem

npm install -g clawdbot completes successfully, but node-llama-cpp main package is not installed. Only platform binaries exist.

Error Message

Local embeddings unavailable.
Reason: optional dependency node-llama-cpp is missing (or failed to install).

Detail: Cannot find package 'node-llama-cpp' imported from 
  ~/.npm-global/lib/node_modules/clawdbot/dist/memory/node-llama.js

node_modules Status

$ ls node_modules/ | grep llama
@node-llama-cpp    # ✅ platform binaries exist
# node-llama-cpp   # ❌ main package missing

Related Issues

• memory_search crashes gateway: node-llama-cpp FileHandle GC error on Node.js 25 #18839 - FileHandle GC error on Node.js 25
• Make node-llama-cpp an optional dependency to reduce install size by ~670MB #17988 - Make node-llama-cpp an optional dependency
• doctor: false positive 'no local model file found' for memory search #28944 - doctor: false positive
• Bug: memory_search tool uses Gemini instead of local despite provider=local (CLI works correctly) #29318 - memory_search uses Gemini instead of local

Workaround

clawdbot gateway config.patch '.agents.defaults.memorySearch.provider = "gemini"'