workspace/skills/catc-inventory/SKILL.md
Catalyst Center device inventory and site management - list/filter devices by hostname, IP, platform, family, role, reachability; view site hierarchy; get interface details per device; device reachability monitoring; cross-reference with pyATS. Use when listing network devices, checking device reachability, auditing software versions, viewing site hierarchy, or finding a device by serial number.
npx skillsauth add automateyournetwork/netclaw catc-inventoryInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
All Catalyst Center tool calls use this invocation pattern:
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 -u $CATC_MCP_SCRIPT
Variable shorthand used throughout this document:
CATC_CMD="CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 -u $CATC_MCP_SCRIPT"
Use the $MCP_CALL protocol handler to invoke MCP tools:
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" TOOL_NAME 'ARGS_JSON'
fetch_devices -- List and Filter Network DevicesFetches devices from the /dna/intent/api/v1/network-device endpoint with extensive filtering. NEVER call this tool without at least one filter parameter -- unfiltered queries can return massive result sets and are blocked by design.
Parameters (all optional, but at least one required):
| Parameter | Type | Description |
|-----------|------|-------------|
| hostname | List[str] | Filter by device hostname(s) |
| managementIpAddress | List[str] | Filter by management IP(s) |
| serialNumber | List[str] | Filter by serial number(s) |
| platformId | List[str] | Filter by platform ID (e.g., C9300-48UXM) |
| family | List[str] | Filter by device family (e.g., Switches and Hubs, Routers, Wireless Controller) |
| type | List[str] | Filter by device type (e.g., Cisco Catalyst 9300 Switch) |
| role | List[str] | Filter by role: ACCESS, DISTRIBUTION, CORE, BORDER ROUTER, UNKNOWN |
| reachabilityStatus | List[str] | Filter by: Reachable, Unreachable |
| collectionStatus | List[str] | Filter by: Managed, Partial Collection Failure, Could Not Synchronize |
| softwareVersion | List[str] | Filter by IOS/NX-OS version string |
| softwareType | List[str] | Filter by: IOS-XE, NX-OS, IOS, AireOS |
| series | List[str] | Filter by product series (e.g., Cisco Catalyst 9300 Series Switches) |
| locationName | List[str] | Filter by full site path (e.g., Global/USA/Building1/Floor1) |
| associatedWlcIp | List[str] | For APs: filter by WLC management IP |
| macAddress | List[str] | Filter by device MAC address(es) |
| id | List[str] | Filter by device UUID(s) |
| roleSource | List[str] | Filter by role source: MANUAL, AUTO |
| offset | str | Starting record for pagination (e.g., "0") |
| limit | str | Maximum records to return (e.g., "50") |
| sortBy | str | Attribute to sort by (e.g., hostname) |
| sortOrder | str | asc or desc |
Response fields per device: id (UUID), hostname, managementIpAddress, macAddress, serialNumber, platformId, softwareVersion, softwareType, type, family, series, role, roleSource, reachabilityStatus, collectionStatus, errorCode, locationName, location, upTime, lastUpdated, lastUpdateTime, associatedWlcIp, apManagerInterfaceIp, and more.
fetch_sites -- List All SitesFetches the complete site hierarchy from Catalyst Center. Takes no arguments.
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_sites '{}'
Response fields per site: id, name, parentId, siteNameHierarchy (full path like Global/USA/NYC/Floor2), type (area, building, floor), address, latitude, longitude.
fetch_interfaces -- Get Interfaces for a DeviceFetches all interfaces for a specific device by its UUID from the /dna/intent/api/v1/interface/network-device/{device_id} endpoint.
Parameters:
device_id (string, required): The UUID of the device (obtained from fetch_devices response id field)CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_interfaces '{"device_id":"a1b2c3d4-e5f6-7890-abcd-ef1234567890"}'
Response fields per interface: id, deviceId, portName, interfaceType, adminStatus, status (operStatus), speed, duplex, ipv4Address, ipv4Mask, macAddress, mediaType, mtu, vlanId, nativeVlanId, description, portMode, lastUpdated, and more.
Identify devices that Catalyst Center cannot reach -- critical for outage detection.
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"reachabilityStatus":["Unreachable"]}'
Analyze the results:
lastUpdated to determine when the device was last seencollectionStatus for additional context (Could Not Synchronize vs Partial Collection Failure)locationName to determine if the outage is site-localizedEscalation: If multiple devices at the same site are unreachable, this likely indicates a site-wide outage (WAN link, upstream switch, power). Escalate to the site-wide outage triage workflow in the catc-troubleshoot skill.
Enumerate all switching devices managed by Catalyst Center.
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Switches and Hubs"]}'
Refine by role:
# Access layer switches only
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Switches and Hubs"],"role":["ACCESS"]}'
# Distribution layer switches
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Switches and Hubs"],"role":["DISTRIBUTION"]}'
# Core switches
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Switches and Hubs"],"role":["CORE"]}'
Refine by platform:
# All Catalyst 9300 series
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"series":["Cisco Catalyst 9300 Series Switches"]}'
# Specific platform ID
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"platformId":["C9300-48UXM"]}'
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Routers"]}'
# Wireless LAN Controllers
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Wireless Controller"]}'
# Unified Access Points
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"family":["Unified AP"]}'
# APs associated with a specific WLC
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"associatedWlcIp":["10.1.100.10"]}'
This is a two-step process: first find the device to obtain its UUID, then query interfaces.
Step 1: Find the device
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"hostname":["CORE-SW-01"]}'
Extract the id field (UUID) from the response.
Step 2: Fetch interfaces using the UUID
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_interfaces '{"device_id":"a1b2c3d4-e5f6-7890-abcd-ef1234567890"}'
Analyze interface data:
adminStatus (UP vs DOWN) and status (operationally UP vs DOWN)adminStatus=UP but status=down -- these are operationally failed interfacesspeed and duplex for mismatchesportModeBuild a report of devices organized by site hierarchy.
Step 1: Retrieve the site hierarchy
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_sites '{}'
Step 2: For each site, retrieve devices at that location
# Devices at a specific site
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"locationName":["Global/USA/NYC/Floor2"]}'
Build the report:
Site Inventory Report
=====================
Catalyst Center: $CCC_HOST
Timestamp: YYYY-MM-DD HH:MM UTC
Global/USA/NYC
Global/USA/NYC/Floor1 (floor)
+--------------+------------------+-----------+------------+--------+
| Hostname | Mgmt IP | Platform | SW Version | Status |
+--------------+------------------+-----------+------------+--------+
| ACC-SW-01 | 10.1.10.1 | C9300-48U | 17.09.04a | Reach |
| ACC-SW-02 | 10.1.10.2 | C9300-48U | 17.09.04a | Reach |
| AP-NYC-F1-01 | 10.1.10.100 | C9120AXI | 17.09.04a | Reach |
+--------------+------------------+-----------+------------+--------+
Global/USA/NYC/Floor2 (floor)
+--------------+------------------+-----------+------------+--------+
| Hostname | Mgmt IP | Platform | SW Version | Status |
+--------------+------------------+-----------+------------+--------+
| ACC-SW-03 | 10.1.20.1 | C9300-24P | 17.09.04a | Reach |
| AP-NYC-F2-01 | 10.1.20.100 | C9130AXI | 17.09.04a | Unrch |
+--------------+------------------+-----------+------------+--------+
Total: 5 devices | 4 reachable | 1 unreachable
Identify devices running non-compliant or mixed firmware versions.
# Find all devices on a specific software version
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"softwareVersion":["17.09.04a"]}'
# Find devices running IOS-XE
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"softwareType":["IOS-XE"]}'
Audit steps:
softwareVersionSoftware Version Audit
======================
Platform: Cisco Catalyst 9300 Series Switches
17.09.04a: 42 devices (APPROVED)
17.06.05 : 3 devices (NON-COMPLIANT -- schedule upgrade)
17.03.04a: 1 device (EOL VERSION -- immediate upgrade required)
Platform: Cisco Catalyst 9800 Wireless Controller
17.09.04a: 2 devices (APPROVED)
# By serial number
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"serialNumber":["FDO2325A0B1"]}'
# By management IP
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"managementIpAddress":["10.1.10.1"]}'
# By MAC address
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"macAddress":["00:1A:2B:3C:4D:5E"]}'
Devices with collection failures may have partial data in Catalyst Center.
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"collectionStatus":["Partial Collection Failure"]}'
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"collectionStatus":["Could Not Synchronize"]}'
Common causes:
When the pyATS testbed is available ($PYATS_MCP_SCRIPT and $PYATS_TESTBED_PATH are set), cross-reference Catalyst Center inventory with live device data for deeper validation.
Compare what Catalyst Center reports vs what the device itself reports:
# Catalyst Center says the device runs 17.09.04a
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_devices '{"hostname":["CORE-SW-01"]}'
# Verify directly from the device
PYATS_TESTBED_PATH=$PYATS_TESTBED_PATH python3 $MCP_CALL "python3 -u $PYATS_MCP_SCRIPT" pyats_run_show_command '{"device_name":"CORE-SW-01","command":"show version"}'
Flag any discrepancy -- this indicates Catalyst Center inventory is stale and needs a resync.
Compare Catalyst Center interface data vs live device state:
# Catalyst Center interface data (may be up to 25 minutes stale)
CCC_HOST=$CCC_HOST CCC_USER=$CCC_USER CCC_PWD=$CCC_PWD python3 $MCP_CALL "python3 -u $CATC_MCP_SCRIPT" fetch_interfaces '{"device_id":"<UUID>"}'
# Live interface state from the device
PYATS_TESTBED_PATH=$PYATS_TESTBED_PATH python3 $MCP_CALL "python3 -u $PYATS_MCP_SCRIPT" pyats_run_show_command '{"device_name":"CORE-SW-01","command":"show ip interface brief"}'
Identify drift: Interfaces that show UP in Catalyst Center but DOWN on the live device (or vice versa) indicate a polling lag or inventory desync.
Always produce a consolidated inventory summary when performing audit workflows:
Catalyst Center Inventory Report
=================================
Controller: $CCC_HOST
Timestamp: YYYY-MM-DD HH:MM UTC
Device Summary
--------------
Total Devices: 156
Switches and Hubs: 92 (59%)
Routers: 12 (8%)
Wireless Controller: 4 (3%)
Unified AP: 48 (31%)
Reachability
------------
Reachable: 152 (97.4%)
Unreachable: 4 (2.6%)
- ACC-SW-15 (10.1.30.15) at Global/USA/NYC/Floor3 -- last seen 2h ago
- AP-CHI-F2-03 (10.2.20.103) at Global/USA/CHI/Floor2 -- last seen 45m ago
- AP-CHI-F2-04 (10.2.20.104) at Global/USA/CHI/Floor2 -- last seen 45m ago
- RTR-WAN-02 (10.0.0.2) at Global/USA/NYC/MDF -- last seen 6h ago
Collection Status
-----------------
Managed: 150
Partial Collection Failure: 4
Could Not Synchronize: 2
Role Distribution
-----------------
ACCESS: 82 | DISTRIBUTION: 8 | CORE: 4 | BORDER ROUTER: 2 | UNKNOWN: 12
Site Summary
------------
Total Sites: 24 (8 areas, 10 buildings, 6 floors)
Sites with unreachable devices: 2 (Global/USA/NYC/Floor3, Global/USA/CHI/Floor2)
Software Version Compliance
----------------------------
IOS-XE 17.09.04a: 140 devices (COMPLIANT)
IOS-XE 17.06.05: 8 devices (NON-COMPLIANT)
AireOS 8.10.185: 4 devices (COMPLIANT)
NX-OS 10.3(2): 4 devices (COMPLIANT)
After completing any inventory operation, record the session in GAIT:
python3 $MCP_CALL "python3 -u $GAIT_MCP_SCRIPT" gait_record_turn '{"input":{"role":"assistant","content":"Catalyst Center inventory audit on $CCC_HOST: 156 devices total, 152 reachable (97.4%), 4 unreachable (ACC-SW-15, AP-CHI-F2-03, AP-CHI-F2-04, RTR-WAN-02). Collection failures: 6 devices. Software compliance: 96% on approved versions. 2 sites with unreachable devices flagged for investigation.","artifacts":[]}}'
testing
Human-in-the-loop escalation via HumanRail — route low-confidence agent decisions, pre-destructive operation approvals, and ambiguous incident tickets to real human engineers. Human answers are verified and returned as structured output. Workers are paid via Lightning Network. Use when the agent is uncertain, when a destructive change needs explicit human sign-off beyond a ServiceNow CR, or when an ambiguous ticket requires human triage before automated handling.
testing
Manage EVE-NG node lifecycle. Use when listing nodes, checking runtime state, creating or deleting nodes, starting or stopping nodes or whole labs, verifying node details, or wiping node NVRAM back to factory defaults.
development
Manage EVE-NG labs and platform inventory. Use when listing labs, checking lab metadata, creating or deleting labs, importing or exporting lab archives, checking EVE-NG health or auth, or verifying available node images before build work.
tools
Execute live CLI commands on running EVE-NG nodes over telnet console. Use when running show commands, making live config changes, verifying protocol state, testing connectivity, checking console readiness, or interacting with IOS, Junos, VPCS, EOS, or NX-OS nodes.