How to use Claude to diagnose ISP peering problems
Claude identified a bad IX-Denver peering connection between the ISP and Fastly; the author later contacted the ISP and says it confirmed the diagnosis.
news.ycombinator.com
Prompt to try
Editorial adaptation, not the source author’s exact prompt. Replace bracketed inputs and review tool permissions first.
Investigate [network symptoms] affecting [service and ISP]. Use available traceroutes, looking glasses and ASN peering information to identify likely routing or peering causes. Explain the evidence and produce a concise diagnosis I can verify with [network provider].
Reach people exploring AI agents
See placements →What happened
Claude identified a bad IX-Denver peering connection between the ISP and Fastly; the author later contacted the ISP and says it confirmed the diagnosis.
Reported by linsomniac on Hacker NewsHow the agent did it
- Asked Claude to diagnose the organisation’s networking problem.
- Ran traceroutes and consulted looking glasses and ASN peering databases.
- Identified a suspected bad peering connection between the ISP and Fastly at IX-Denver.
Where a person stepped in
The author assessed the diagnosis against known peering arrangements and contacted the ISP, which confirmed it.
A person guided the agent closely as it worked.
During a networking problem, the author's coworker asked Claude to diagnose the issue while the author was switching to a mobile hotspot to determine whether the problem was local or external. Claude investigated with traceroutes, looking glasses and ASN peering databases.
It identified a bad peering connection at IX-Denver between the organisation's ISP and Fastly, and said the ISP needed to withdraw the advertisement. The author found this plausible because they already knew both networks peered at that exchange. That night, the author contacted the ISP to ask whether this had happened, and the ISP confirmed it.
The same post describes broader infrastructure work with Claude. It had updated Ansible YAML controlling Kafka and was building a script for the rolling update process. It also helped manage Nix configuration and packaged the Git Butler CLI. The author periodically asked it to review production syslog for security problems and five actionable improvements; that review identified Kafka configuration changes related to the rolling update work. The networking diagnosis is the case with external confirmation in the account.
Who it's for: infrastructure operators investigating network symptoms who can verify a diagnosis with the responsible provider.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Hacker News · linsomniac
Frequently asked questions
What evidence did Claude consult?
The post reports traceroutes, looking glasses and ASN peering databases during the networking investigation.
How was the diagnosis verified?
The author already knew the ISP and Fastly peered at IX-Denver. They contacted the ISP that night, and report that it confirmed the diagnosis.
More details
Remaining steps
- Contacted the ISP that night and received confirmation of the diagnosis.
About this account
Claude investigated network paths and peering information and surfaced a specific suspected provider-side cause.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=47806419. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: linsomniac on Hacker News
Source published 17 Apr 2026
Source checked 14 Oct 2026
Tools and topics
Claude · Traceroute · Looking glasses · ASN peering databases
Networking · ISP peering · Troubleshooting
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.