HPE Networking Assistant
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MIST_REGION | No | Optional override for the Mist region (e.g., 'global', 'eu', 'ap'). If not provided, the region is auto-detected. | |
| MIST_API_TOKEN | Yes | Your Mist API token. Required for authentication. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| start_setupA | Run first-run onboarding: auto-detect the Mist region from the token, discover organizations and sites, run validation, and return a READY FOR USE report. Call this when the user is setting up or asks to get started. Optionally pass organization="" to choose among multiple orgs. |
| get_statusA | Report the current mode (read-only vs read-write), region, organization, and whether the token can make changes. Use when the user asks about their mode or write access. |
| get_organizationsA | List the Juniper Mist organizations the API token can access. |
| get_sitesC | List the sites in an organization. |
| get_access_pointsB | List access points (APs) in an organization, optionally filtered to a site. |
| get_switchesA | List switches in an organization, optionally filtered to a site. |
| get_clientsA | List currently connected wireless clients (org-wide or for one site). |
| get_offline_access_pointsB | Report access points that are currently offline (disconnected). |
| generate_health_reportA | Generate a network health report as Markdown: device totals, online/offline counts, offline AP list, and per-site breakdown. The user can then save it as a file. Set include_clients=false to skip the (slower) org-wide client count. |
| generate_inventory_reportA | Generate a full device inventory report as Markdown (every AP and switch with model, serial, MAC, site, status, and firmware version). |
| generate_firmware_reportA | Firmware compliance report (Markdown): for each AP/switch model, the version most of the fleet runs and which devices are behind it. Use to find firmware drift. |
| generate_topologyA | Build a per-site network topology as a Mermaid diagram (plus structured nodes/edges): gateways, switches, and APs linked via switch-port LLDP neighbors. Provide a site name; set include_clients=true to add connected clients. Render the Mermaid or save it. |
| find_clientA | Locate a wireless client by MAC address or hostname: which site, AP, and SSID it is on, plus IP, band, and signal. Provide mac or hostname. |
| trace_clientA | Trace a wireless client's recent connection events (association, auth, DHCP, roam, disconnect) to troubleshoot why it can't connect or roams poorly. Returns an event timeline, type counts, and highlighted failures. |
| get_nac_clientsA | List Access Assurance (NAC) clients authenticated to the network — user, client type (wired/wireless), auth type (EAP-TLS/PEAP/MAB), SSID, VLAN, matched auth rule, and status. Optional filters: mac, auth_type, type. |
| troubleshoot_authenticationA | Troubleshoot 802.1X/MAB authentication using Access Assurance (NAC) events: returns the auth-event timeline, per-type counts, and highlighted failures (denied/rejected/errored). Focus on one identity with mac (client MAC) and/or user (username or certificate CN). |
| generate_nac_dashboardA | Build a self-contained HTML Access Assurance (NAC) dashboard: summary cards and bar charts for auth types, client types, status, event types, and top failing users/rules. Returns HTML the user can save as a .html file and open in a browser. |
| get_marvis_actionsA | List Marvis (AI) suggested actions — Mist's prioritized view of what's wrong (offline switches/APs, non-compliant firmware, misconfigured ports, RF/DFS issues) and the recommended fix. Use for 'what's wrong with my network?'. Pass status="open" for active items only. |
| get_alarmsA | List organization alarms over a window with per-severity and per-type counts (device/switch/gateway offline, restarts, security, Marvis). Optional severity filter. |
| get_wired_clientsA | List wired clients (devices seen on switch ports): switch MAC, port, VLAN, IP, and vendor. Optional mac / hostname / site_id filters. Complements get_clients (wireless). |
| get_sleA | Service Level Expectations (SLE) per site — Mist's user-experience scores as percentages (e.g. ap-health, coverage, capacity, successful-connects, switch/gateway/wan health). Use to gauge experience quality by site. |
| get_switch_portsA | Switch/device port statistics: link state, speed/duplex, PoE on/off and draw, LLDP neighbor, and traffic. Filter by switch_mac, site_id, or up (true/false). |
| export_org_configA | Export a read-only backup of the org configuration as JSON (org settings, sites, networks, VPNs, templates, WLANs, NAC rules, webhooks, etc.) for the user to save to a .json file. Secrets are excluded/masked. |
| set_active_orgA | Set the default organization for this session by name or id. Useful when the token can access several organizations — subsequent tools then target the chosen org. |
| diff_org_configA | Compare the current org configuration against a saved backup file (config drift): reports added, removed, and changed objects per resource type. baseline_file is a path to a JSON file previously produced by export_org_config. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 25 tools
Each tool targets a distinct resource and action, from device inventory and client lookup to authentication troubleshooting and config diffing. Even seemingly similar tools like get_clients, get_wired_clients, and get_nac_clients are clearly differentiated by their descriptions (wireless vs. wired vs. Access Assurance). The only mild overlap is between get_access_points and get_offline_access_points, but the latter is a specific filtered view, so confusion is unlikely.
All tool names follow a consistent verb_noun snake_case pattern. The verbs vary semantically (generate, get, find, trace, troubleshoot, start, set, export, diff) but are predictable for their function: generate_ for reports, get_ for lists/retrievals, and so on. This consistency allows an agent to infer behavior from the name alone.
With 25 tools, the server sits at the upper boundary of the 'heavy' range (16–25). The domain is broad, covering inventory, clients, NAC, Marvis, config management, and session setup, so many tools are justified. However, some report generators (inventory, firmware, health, NAC dashboard) could potentially be consolidated, making the count feel slightly bloated.
The tool surface provides thorough coverage for network observability, troubleshooting, and configuration backup/compare. It includes inventory, health, client tracking, authentication analysis, Marvis insights, alarms, SLE scores, and port-level details. The main gaps are the lack of write operations (e.g., implementing Marvis fixes) and no singular device detail lookup, but these feel intentional given the assistant's likely read-only role.