Endpoints - dubsector/meraki-api-emulator GitHub Wiki

All live under /api/v1. ENDPOINTS.md lists each operation with its method and path. In short:

  • Organizations: networks, devices, statuses and availability, power supplies, uplinks and uplink addresses, memory use over time, syslog servers and roles by network, VPN, clients and client search, client bandwidth over time, top-N summaries (applications and their categories, clients and manufacturers, devices and models, SSIDs, networks by status, appliances by utilization, switches by energy), switch PoE power over time, admins, login security, SAML SSO with its IdPs and roles, SNMP access, Dashboard branding policies and their priorities, early access features and opt-ins, splash themes and their assets, Insight applications and monitored media servers, AP auto locate devices and statuses, provisioning pipeline jobs, cloning, the OpenAPI document, licenses with their assignments and moves, co-term licenses and their moves, network groups and their overview, network moves and combining networks, policy objects and policy object groups, organization-wide firewall rulesets and rules with their application categories, organization-wide group policies with their adaptive policy group, appliance VLAN and firewall ruleset assignments, policy assignments by client, adaptive policy groups, ACLs, policies, settings and overview, config templates with their switch profiles and ports, packet captures and their schedules, controller migrations, cellular data profiles and the cellular device views, inventory with claiming and releasing, order previews and claims, its end-of-life overview and device swaps, provisioning statuses, the integrations it can deploy and has deployed, XDR per network, the Cisco Spaces link and removing it, action batches, the change log and the API request log.
  • Networks: binding to config templates, traffic shaping application categories and DSCP tagging options, splitting combined networks, devices and claiming or removing them, clients with their daily usage, application usage, policies (also listed by client) and splash authorization, Meraki authentication users (802.1X, splash guest and client VPN), splash login attempts, NetFlow and traffic analysis settings, VLAN profiles with their assignments by device, events and event types, traffic, settings, syslog, SNMP, alert settings, webhook servers, payload templates and tests, group policies, alert history, firmware upgrades (also across the organization, per upgrade and per device), rollbacks and staged switch upgrades (groups, their order and events), floor plans with their devices and AP auto locate jobs, Insight application health, and link layer topology.
  • Alerts and webhooks: assurance alerts across the organization with overviews by network, by type and over time, dismissing and restoring them, alert profiles and the alert taxonomy, organization-wide alert configs, webhook alert types, the webhook log and callback statuses, plus per-network health alerts.
  • Security appliance (MX): LAN ports, VLANs, L3, L7 and cellular firewall rules and the L7 application categories, firewall spoofing protection, port forwarding, 1:1 and 1:many NAT, static multicast forwarding and uplink NAT (also across the organization), connectivity monitoring destinations, warm spare and swapping it (also as redundancy settings, and across the organization), Wi-Fi on MX models with a radio (radio settings, RF profiles and the four MX SSIDs), vMX authentication tokens, Cisco Secure Routers (DOM readings of their optics over time and packet counters per interface across the organization), static routes, site-to-site VPN and hub BGP, the organization's third-party VPN peers, IPsec SLA policies and VPN firewall rules, content filtering and its categories, intrusion and malware settings (and the organization's allowed intrusion rules), traffic shaping (global and per-uplink bandwidth limits, rules, uplink selection, custom performance classes and VPN exclusions, also across the organization), SD-WAN internet policies, security events per network, client and organization, an uplink status overview, DHCP subnets and uplink settings, the organization's local DNS profiles and records and split DNS profiles with their network assignments, wired L3 interfaces (also across the organization), static delegated IPv6 prefixes and the prefixes IPv6 VLANs take from them, the organization's VRF setting, every appliance's port configuration across the organization and updating a port by device, plus Secure Access (SASE): the organization's integration, the MX networks and templates it can take, regions and their connectors, and attached sites with their default route setting and connectivity now and over time. Each MX network can also connect a Cisco Umbrella account, apply and remove its DNS policies, set domain exclusions and turn protection on or off.
  • Switches (MS): port config and live status per switch and across the organization, organization-wide port counts by media and speed, clients per port, LLDP and CDP discovery and port usage history, the DHCP servers a network sees, packet counters, port cycling, LLDP and CDP neighbors, stacks and layer 3 switches with their interfaces, DHCP and static routes, OSPF, multicast settings and rendezvous points, ACLs, access policies, QoS rules and their order, DSCP to CoS mappings, STP, MTU, storm control, the alternate management interface, link aggregations, warm spare, port schedules, adaptive policy groups and peer SGT on ports, the DHCP server policy with ARP inspection trusted servers and warnings, and cloning one switch's settings onto others.
  • Wireless (MR): SSIDs (with their adaptive policy group) and their firewall, splash, Bonjour forwarding, device type policy, EAP timer, Hotspot 2.0, OpenRoaming, outage schedule, traffic shaping and VPN settings, identity PSKs, RF profiles, radio settings, overrides and status, AutoRF settings, Air Marshal scan results, rules and default policy, Bluetooth settings per network and AP, Bluetooth clients, billing plans, location scanning, the alternate management interface (with per-AP IPv6 addresses), client counts, usage, connection and latency stats (also per client), client connectivity events and latency history, data rate and latency over time, failed connections, channel utilization (also per radio in network health), signal quality and mesh statuses, plus organization-wide client counts per AP, clients impacted by connection failures per SSID, client usage by network and SSID, packet loss by client, AP and network, channel utilization over time, Ethernet and PoE status, power mode and CPU load history, BSSID statuses, APs impacted by outages, radio overrides and RF profile assignments by AP, AutoRF, Air Marshal, location scanning, MQTT and OpenRoaming settings by network, scanning API receivers, Air Marshal rules, AutoRF channel recalculation, SSID profiles with their assignments to SSIDs (also by network), the L2 isolation allowlist, AP port profiles with assigning APs and setting the default, zero touch deployments, RadSec certificate authorities with their CRLs and delta CRLs, Zigbee settings per network with gateways, door locks and their enrollment and disenrollment jobs, and electronic shelf label settings per network and AP.
  • Cameras (MV): quality and retention, video and MV Sense settings per camera, object detection models, video links, snapshots and clips, wireless profiles per camera, plus each network's quality retention profiles, camera wireless profiles, recording schedules and MQTT brokers (which wireless networks have too), the organization's camera permission scopes and roles, area and crossing line boundaries with people and vehicle counts, custom analytics artifacts and each camera's custom analytics, and camera onboarding statuses.
  • Sensors (MT): readings history and the latest reading per sensor across the organization, sensor to gateway connections, each network's alert profiles with the alerting sensors now and alert counts over time by metric, commands (downstream power on an MT40 and refreshing readings) with their history, livestream roles between sensors and cameras per device and per network, and which MQTT brokers get sensor data.
  • Cellular gateways (MG): SIM and APN settings, band masks, geolocation, LAN settings and port forwarding rules per gateway, each network's uplink bandwidth limits, DHCP, connectivity monitoring destinations and subnet pool, the organization's cellular uplink statuses, and its eSIMs (inventory and status, service providers, provider accounts with their communication and rate plans, and profile swaps). The organization cellular views (devices for data profiles, data usage now and over time, geolocations, bands and towers) list the gateways.
  • Campus gateways: clusters (created, updated, provisioned with a failover target and deleted per network) and their list across the organization, failover targets and the clusters that can back each one up, the networks and SSIDs tunneling through them with their counts, the clusters a wireless network can tunnel to, tunnel encryption settings per network and cluster with a batch update, per SSID mDNS gateway rules, the APs tunneling through them (connections) and their up and down counts, client usage by network and cluster, and local uplink overrides.
  • Wireless LAN controllers: Catalyst 9800 controllers monitored from the cloud: their overview (clients, joined APs, redundancy role and firmware), the Catalyst APs joined to them and their connections, layer 2 and layer 3 interfaces with their status changes and usage, packet counters, interface traffic and CPU use over time, client counts over time, SSO redundancy statuses and failover history, and availability changes.
  • Systems Manager (SM): enrolled devices with their optional fields and filters, and per device its certificates, check-in history, desktop logs and performance history (laptops), cellular use (phones and cellular iPads), command log, installed profiles, network adapters, restrictions, security centers (Windows), software and saved Wi-Fi networks; the network's profiles, and its owners with their profiles and software. Devices can be checked in, locked, rebooted, shut down, wiped, unenrolled, renamed, retagged and moved to another SM network, apps installed and uninstalled, and supervised iPhones and iPads sent an activation lock bypass. Target groups keep a tag scope. Each SM network lists its trusted access configs and the owners' devices that use them, which can be deleted. Organizations have limited access roles, an APNS certificate, a VPP account and Sentry policies applying a network's group policies to SM devices by tag. Any network answers PII lookups (keys, SM devices and owners for a MAC, email, username, serial or IMEI) and takes delete and restrict processing requests.
  • Identity: the admin every API key acts as, the keys used this run, and generating and revoking keys.
  • Subscription licensing: the purchasable entitlements, subscriptions across organizations with their seats and filters, claiming one by key (or checking a key first), binding networks to it, and its compliance status.
  • Devices: device details, clients, loss and latency history, MX performance and management interface, reboot, and the live tools (ping from and to the device, ARP and MAC tables, cable test, LED blink, wake on LAN, port cycle and port status, power usage, multicast routing, throughput test, and routing table lookups and summaries on Secure Routers) with API callbacks.