antoninolorenzo/ai-ops/network-traffic-analysis
Claude Code and agent skill by antoninolorenzo/ai-ops
- community
Checking that the SKILL.md file is still in the repository.
Capture and analyze network traffic to extract credentials, session tokens, flags, or other sensitive data in transit. Load when the objective involves intercepting traffic between hosts or observing network communications.
Install
npx vskill@latest install antoninolorenzo/ai-ops/network-traffic-analysisSKILL.md
Network Traffic Analysis
The general process is always: capture traffic to a file → analyze the file. Never stream raw packet output to the terminal, it will flood context immediately.
Two capture modes exist depending on your position relative to the traffic:
- Passive: you are already in the traffic path (traffic is destined to you, or it is broadcast on the segment). Use
tcpdumpto capture. - Active (MITM): traffic flows between two other hosts and you need to insert yourself into the path. Requires concurrent redirection + capture; use a Python script.
Understand the Network Layout
Identify: your interface name, your IP, the local subnet, and the gateway IP. These are needed for all subsequent steps.
Discover What Is on the Wire
Run a short broad capture to understand what traffic exists before committing to a targeted approach. Limit by packet count to avoid large files.
To analyze the capture suppress per-packet output, show only summaries:
From this output determine:
- What protocols are in use (TCP, UDP, ICMP, ARP, HTTP, etc.)
- Which pairs of hosts are communicating
- Whether you are already in the traffic path or not
Decide which capture mode to use:
If the traffic you need to observe is already reaching your interface (it is destined to you, or it is broadcast), passive capture is sufficient.
If the traffic flows between two other hosts and you need to intercept it, you must insert yourself into the path. How you insert yourself depends on the network topology: identify where the hosts are relative to you (same L2 segment, different subnet, DNS-dependent, etc.) and choose the appropriate redirection technique accordingly.
If the discovery capture is empty, re-examine the task description, the hosts may not be communicating, or the interface or subnet may be wrong. Try setting tcpdump interface to another interface or any interface (-i any).
Passive Capture
Capture with a targeted BPF filter to limit file size and noise:
BPF filter primitives: host, net, port, tcp, udp, arp, icmp. Combine with and, or, not.
MITM Capture
When traffic flows between two other hosts, you need to redirect it through your machine while forwarding it so the hosts stay connected, and capture it in transit. The appropriate redirection technique depends on the network topology. Determine this from the layout and from the task description, then implement accordingly.
For whichever technique you use, the capture pattern is the same: run redirection and capture concurrently in a single script using scapy, write to a pcap file, stop automatically after a fixed duration. This is necessary because the terminal is non-interactive, so backgrounded processes cannot be managed across separate commands reliably.
Here is an example of how to write packets to a .pcap file using AsyncSniffer and PcapWriter in scapy:
Verify the capture has content before analyzing:
If the capture is empty or only shows ARP traffic, the redirection did not succeed. Check that both IPs are reachable and that the correct gateway was identified.
Analyze the Capture
The analysis step is identical regardless of how the capture was obtained.
Identify protocol and conversation structure first
Extract payload content
Match the extraction command to the protocol identified above:
Broad sweep when protocol is unknown or above filters return nothing
If you have a whiteboard and found the target data, record it:
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingai_ops/core/tools/load_skill/bundled/network-traffic-analysis/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/antoninolorenzo/ai-ops/network-traffic-analysis)