Network troubleshooting
Triage incidents, rebuild outage timelines, read captures, and isolate fault domains fast.
- Incident Triage Analyst
- Outage Timeline Builder
- Packet Capture Analyst
Six stages, 26 guided skills. Work the way experienced teams work—triage, change, capacity, standards, and reporting—with human approval on every critical decision.
Prototype · Stripe checkout and delivery connection pending
Built for accountable network work
Evidence first · Human approval · Repeatable outputs
› /network-triage
Loading Incident Triage Analyst skill…
✓ Symptom timeline normalized
✓ Fault domain isolated to WAN edge
! Config change awaiting human approval
Six stages, worked in order. Each one teaches a piece of how a network team operates, then hands you the skills to do it on your own network.
Before you change anything, learn to describe what you actually run—dependencies, current state, and how the desk hands off work.
The core reflex of any network team: take a vague symptom, isolate the fault domain, and know when to escalate.
What separates a strong team from a busy one: turning every incident into a written cause, a ranked fix list, and a habit that sticks.
Learn the discipline that lets teams touch production calmly—blast radius, maintenance windows, exceptions, and early warning.
Move from reacting to governing. Write the rules your network runs by, then show your work to auditors and vendors.
The stage most engineers skip. Turn what you found into language that wins budget, trust, and the next project.
Once you have worked the path, this is your reference shelf. Every skill is a task-specific procedure—add your topology, standards, and telemetry to get consistent drafts and clear decision support.
Triage incidents, rebuild outage timelines, read captures, and isolate fault domains fast.
Assess blast radius, plan maintenance windows, and review exceptions before production.
Cut alert noise, form testable hypotheses, and plan capacity well ahead of demand.
Draft clear standards, map configs to requirements, and keep a living risk register.
Turn technical findings into clear recommendations for leadership and clients.
Document runbooks, escalation paths, dependency maps, and clean handoffs.
Supply topology, standards, tool exports, and the output you need.
A bounded workflow with clear assumptions, evidence rules, and stop conditions.
Check sources, affected devices, impact, confidence, and next steps.
A qualified engineer authorizes every change, failover, and maintenance window.
A portable operating layer built for Claude Code, Codex, and other file-aware AI workspaces.
✓ 25+ task-specific network skills
✓ Workspace instruction template
✓ Evidence and citation standards
✓ Human-approval guardrail template
✓ Example inputs and outputs
✓ Setup and customization guide
No. It organizes and accelerates defined work. Qualified engineers stay responsible for validation, authorization, changes, and business decisions.
Any file-aware AI workspace, such as Claude Code and Codex. Results vary by model, tools, context, and configuration.
Not on their own. Access depends on the tools you configure. Start read-only, apply least privilege, and require human approval for every change.
No. The templates support analysis, planning, and drafting. They are not certification, legal advice, or a guarantee of availability, performance, or compliance.
Start with focused skills. Tune them to your environment. Keep your engineers in control.
Prototype notice: checkout is not connected yet.