#general
With github.com/archestra-ai/archestra/pull/6488 was the intention on this to hard block all access to the following ranges from MCP pods with no way to override if the environments networkPolicy is set to unrestricted (Allow All)?:
-
10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16 - RFC 1918 private ranges (cluster pods, services, nodes) <--- This is the primary one I'm wondering about -
169.254.0.0/16 - Link-local / cloud metadata endpoints (AWS IMDSv1, GCP, Azure)168.63.129.16/32- Azure platform metadata (a public IP outside the private ranges)
-
100.64.0.0/10 - Carrier-grade NAT (RFC 6598)--
127.0.0.0/8 - Loopback--
0.0.0.0/32 - Treated as localhost by some HTTP libraries
Hi i have been testing out archestra to see if it would fit our companies needs, but i have a couple of question i cant figure out.
1. When i create an mcp registry and attach it to a gateway is there a way of having for each user it's own "identity"/"credentials" that the mcp from the registry could connect to or do i need to register for each user its own mcp (example is linear mcp). The goal is that if user X uses the mcp gateway it will use the registered linear mcp with his credentials instead of a shared one. Would we register for each user it's own MCP of linear?
2. Is there anything on the roadmap for having more layers than organizational - team - personal? We would love to have nested teams/projects
👋2
6 replies
hello, currently experimenting archestra for our organization as centralized gateway for agent, mcp and llm proxies.
We are using Vertex AI and wondering how to make archestra to list Sonnet 5 in model choice?
because using our mounted Vertex SA, archestra only shows all Gemini model.
Using the same SA and testing it locally using curl, it does respond correctly so i believe it should not be the settings from model garden or access issue.
--impersonate-service-account="$SA")
curl -i -X POST \
"<https://aiplatform.googleapis.com/v1/projects/company/locations/global/publishers/anthropic/models/claude-sonnet-5:rawPredict>" \
-H "Authorization: Bearer $TOKEN" \
-H "Content-Type: application/json" \
-d '{
"anthropic_version": "vertex-2023-10-16",
"max_tokens": 32,
"messages": [
{
"role": "user",
"content": "Reply only with OK"
}
]
}'
WARNING: This command is using service account impersonation. All API calls will be executed as [platform-eng-vertex-ai-service@company.iam.gserviceaccount.com].
HTTP/2 200
.
.
.
{"model":"claude-sonnet-5","id":"msg_vrtx_011Cek8kRJKHbBqjfLRmyT9H","type":"message","role":"assistant","content":[{"type":"text","text":"OK"}],"stop_reason":"end_turn","stop_sequence":null,"stop_details":null,"usage":{"input_tokens":13,"cache_creation_input_tokens":0,"cache_read_input_tokens":0,"cache_creation":{"ephemeral_5m_input_tokens":0,"ephemeral_1h_input_tokens":0},"output_tokens":4,"output_tokens_details":{"thinking_tokens":0}}}%```👀1
7 repliesRead-only live mirror of Archestra.AI Slack
👋Join the discussion withAI enthusiasts!