📋 Pre-flight Checks
📄 Which Document?
README.md
🔍 What's Wrong or Missing?
The exact steps to setup engram aren't clear to me. I assumed that the engram setup x would be sufficient but my agent doesn't seem to agree. Also, it's unclear if I now have to engram serve or engram mcp after each boot. Then I also see some "add X to your agent's Y" across the docs, but then again stating this was the "nuclear option".
Tbh, maybe I'm a bit dumb today - but the setup doesn't seem to just work.
Also, I have agents running in docker that I'd like to connect to engram MCP on the host. But that doesn't seem to be possible? I only see stdio MCP. But what's ENGRAM_URL for? And why do the docs state
ENGRAM_URL=http://127.0.0.1:7437 pi
That's not a valid URL or valid shell syntax. Is the pi accidental? Was /pi intended?
💡 Suggested Improvement
No response
📎 Additional Context
No response
📋 Pre-flight Checks
📄 Which Document?
README.md
🔍 What's Wrong or Missing?
The exact steps to setup engram aren't clear to me. I assumed that the
engram setup xwould be sufficient but my agent doesn't seem to agree. Also, it's unclear if I now have toengram serveorengram mcpafter each boot. Then I also see some "add X to your agent's Y" across the docs, but then again stating this was the "nuclear option".Tbh, maybe I'm a bit dumb today - but the setup doesn't seem to just work.
Also, I have agents running in docker that I'd like to connect to engram MCP on the host. But that doesn't seem to be possible? I only see stdio MCP. But what's
ENGRAM_URLfor? And why do the docs stateThat's not a valid URL or valid shell syntax. Is the
piaccidental? Was/piintended?💡 Suggested Improvement
No response
📎 Additional Context
No response