Replies: 1 comment
|
I've answered the questions on Discourse: https://discourse.margo.org/t/governed-mcp-based-ot-diagnostics-as-a-margo-edge-application-reference-feedback/43/2?u=philip_presson_abb This is where we'd like to keep public communication, so I'm closing this discussion thread. Please post follow-up questions on Discourse. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
A spec question on the Application Description (
margo.org/v1-alpha1), grounded in a concrete candidate app.Industrial-AIOps (
iaiops, MIT, on PyPI + the MCP Registry) is a headless, pure-Python OT diagnostics workload that runs as a governed MCP server — a vendor-neutral, read-first tap across ~14 field protocols (OPC-UA, Modbus, S7comm, EtherNet/IP, PROFINET-DCP, BACnet/IP, MTConnect, MQTT-Sparkplug, HART-IP, …) plus a cross-protocol RCA "brain", all exposed as MCP tools with audit / budget / risk-tier / undo governance. A site enables one profile selecting which connectors load.I've drafted a
margo.yamlto thev1-alpha1schema — shape below:Three questions:
profileparameter (SelectValidationSchema →IAIOPS_MCPenv var). Is that the intended idiom, or would you expect N app packages (one per profile)?v1-alpha1seems to have no secret-typed parameter; how should a workload declare a secret param, plus "read-first, outbound-only to configured OT endpoints, no inbound"? Governance is core to this app.Happy to open-source the full
margo.yaml+ a hardened container recipe as a worked example for a "governed, read-only OT tool app." Thanks!All reactions