Cisco network ops

Skill olandodeflexy/cisco-network-ops-skill/skills/cisco-network-ops

Use when reviewing Cisco network changes, incidents, configs, logs, or automation for risk, evidence, validation, and rollback.From its SKILL.md

Install

npx -y skills add olandodeflexy/cisco-network-ops-skill --skill cisco-network-ops

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as Apache-2.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

7.3 KB, as published. Nobody here has run it

Cisco Network Operations

Diagnose-first guidance for Cisco network operations. Use offline artifacts first: configs, diffs, CLI output, parsed telemetry, NMS alerts, topology notes, tickets, and automation code.

This skill is advisory by default. It does not assume direct access to routers, switches, firewalls, controllers, or production networks.

Reference files live under references/ relative to this SKILL.md; load them from the skill directory, not from the caller's current working directory.

Response Contract

For safety-critical Cisco network operations responses, include:

Assumptions - platform, OS version if known, topology role, environment criticality, access path, and missing context.
Risk category - one or more categories from the diagnose table.
Evidence - exact config, diff, log, command output, alert, or automation behavior used.
Recommendation - next diagnostic step or remediation, with tradeoffs.
Pre-checks - read-only commands, backups, neighbor state, interface state, route/policy evidence.
Execution guidance - ordered steps or draft config only when appropriate.
Post-checks - verification commands, parsed checks, telemetry, or before/after comparisons.
Rollback - inverse config, checkpoint/restore path, abort criteria, and retained evidence.
Escalation notes - when to involve TAC/vendor/internal escalation and what data to collect.

Small factual answers may compress the contract. Production-impacting advice must keep the full shape.

Workflow

Capture context: platform, OS version, device role, topology, maintenance window, criticality, change intent, and available artifacts.
Classify the failure mode using the diagnose table.
Load only the reference files matching the category.
Separate evidence from inference. State when a conclusion is unproven.
Prefer read-only diagnostics before any configuration.
For state-changing work, produce pre-checks, execution steps, post-checks, rollback, and abort criteria.
Validate with lab, parser, pyATS/Genie, Batfish, lint, or staged post-checks where applicable.
Emit the Response Contract. For an end-to-end pattern, see Example Response.

Diagnose Before You Generate

Failure category	Symptoms	Primary references
Change blast radius	Large target set, unclear affected devices, no window, no rollback, no pre/post evidence	Change Safety, Testing and Validation
Platform mismatch	IOS XE/NX-OS/IOS XR syntax mixed, unsupported commands, wrong commit/rollback model	Platform Matrix, Quick Reference
Config drift / idempotency	Running config differs from intended state, repeated automation changes, unbounded CLI pushes	Automation and APIs, Testing and Validation
Routing convergence risk	BGP/OSPF/EIGRP/IS-IS adjacency changes, route loss, redistribution, policy edits	Routing Troubleshooting, Change Safety
Layer 2 risk	STP, VLAN, trunk, port-channel, vPC, loop, blackhole, MAC move symptoms	Switching Troubleshooting, Change Safety
HA / redundancy risk	HSRP/VRRP/GLBP, vPC, VSS, StackWise, dual supervisor/RP state, asymmetric failover	Switching Troubleshooting, Platform Matrix
ACL/NAT/policy shadowing	Rule order errors, broad permits/denies, NAT precedence, object-group mistakes	Security Policy, ASA/FTD Operations
Secret exposure	Passwords, SNMP communities, keys, tokens, configs or logs exposing credentials	Security Policy, Automation and APIs
Observability gaps	Missing before/after evidence, weak telemetry, no alert correlation, unclear success signal	Telemetry and Observability, Quick Reference
Validation blind spots	No lab, parser, dry-run, pyATS, Batfish, lint, or staged verification	Testing and Validation

When to Use This Skill

Activate for Cisco network operations tasks: config review, planned change risk analysis, incident triage, routing/switching troubleshooting, ACL/NAT review, HA checks, telemetry interpretation, and Cisco automation review.

Also activate for automation touching Cisco devices through Ansible, Nornir, Netmiko, pyATS/Genie, NETCONF, RESTCONF, controller APIs, or streaming telemetry.

Do Not Use This Skill For

General networking definitions that do not involve Cisco operations.
Certification tutoring or exam-prep drills.
Official Cisco/TAC authority claims.
Direct production device access, credential handling, or autonomous remediation.
Full SD-WAN, ACI, wireless RF, UCS, Intersight, or Webex design.

Platform Scope

Platform	v1 behavior
IOS XE	Primary: change review, troubleshooting, command guidance, automation review
NX-OS	Primary: change review, troubleshooting, command guidance, automation review
IOS XR	Primary: change review, troubleshooting, command guidance, automation review
ASA/FTD	Secondary: ACL/NAT/policy/routing triage; no full firewall design or deployment modeling
ACI, Catalyst Center, Meraki, SD-WAN	Secondary: operational context and escalation guidance only

When platform or version is missing, say so and avoid platform-specific commands that may be wrong.

Safety Rules

Treat all production config as draft text until a human reviews it.
Label commands as read-only or state-changing when giving runbooks.
Never imply the agent has live device access unless a tool explicitly provides it.
Prefer show/parse/compare evidence before configure terminal, commit, write memory, reload, clear, shut/no shut, or failover actions.
Do not recommend direct production writes without maintenance context, pre-checks, post-checks, rollback, and abort criteria.
Redact secrets from pasted configs and avoid repeating credentials in responses.
For ambiguous symptoms, give the next evidence-gathering command before giving a fix.

Quick Decision Matrix

User intent	Default output
"Review this change"	Risk categories, evidence, pre-checks, post-checks, rollback, approval questions
"Troubleshoot this output"	Most likely causes, confidence, next read-only commands, escalation data
"Generate config"	Ask for missing platform/topology; emit draft config plus validation and rollback
"Review automation"	Idempotency, blast radius, secret handling, parser/test strategy
"Summarize incident"	Timeline, impact, evidence, current hypothesis, next actions, escalation bundle