Works with the AI Agents you already use
Network Threat Hunter
Detect C2 beacons, DNS tunneling, and data exfiltration by scoring network telemetry for regularity and entropy.
$1.49
Network Threat Hunter
Example session with this skill installed
Analyze this Zeek conn.log for any signs of beaconing or C2 activity. Highlight the top 5 suspicious destinations based on regularity.
- Read your context and instructions
- Compiled the network threat hunter
I have processed the provided Zeek conn.log file. The analysis calculated interval regularity, payload size consistency, and connection volume to identify potential C2 beacons.
Network Threat Analysis Results
| Source IP | Destination IP | Dest Port | Conn Count | Regularity Score | Size Score | Findings |
|---|---|---|---|---|---|---|
| 192.168.1.45 | 91.228.167.21 | 443 | 842 | 0.942 | 0.981 | High-confidence beacon; consistent 60s interval. |
| 192.168.1.12 | 104.26.10.233 | 443 | 1204 | 0.895 | 0.412 | Periodic heartbeat; payload size varies significantly. |
| 192.168.1.45 | 185.199.110.153 | 80 | 45 | 0.821 | 0.884 | Periodic check-in; potential T1071 mapping. |
| 172.16.0.5 | 52.114.32.12 | 443 | 2150 | 0.710 | 0.120 | Noisy regularity; likely legitimate telemetry. |
| 192.168.1.102 | 203.0.113.5 | 53 | 540 | 0.680 | 0.950 | Potential DNS tunneling; check dns.log for entropy. |
Analysis Summary
The destination 91.228.167.21 shows a near-perfect regularity score (0.942) with extremely low jitter. This behavior is highly characteristic of a Cobalt Strike or similar C2 agent configured with a 60-second sleep and 0% jitter. The connection at 185.199.110.153 also warrants investigation due to the combination of interval and size regularity.
Next steps
- Run
scripts/beacon_hunt.py --pair 192.168.1.45,91.228.167.21,443to extract raw timestamps for visualization. - Cross-reference
91.228.167.21against EDR logs to identify the specific process initiating these calls. - Check threat intelligence feeds for the destination IP to confirm known C2 infrastructure.
Connects securely to your tools. The creator never sees your data.
About this skill
Security logs are noisy, making it nearly impossible to spot a single command-and-control beacon or DNS tunnel manually. This skill analyzes raw network telemetry to identify stealthy outbound threats that bypass traditional signature-based detection.
What it does
- Beacon detection identifies periodic connections by calculating interval regularity and jitter scores.
- DNS tunneling analysis flags high-entropy labels, long query names, and unusual subdomain cardinality.
- Exfiltration tracking surfaces high-volume outbound transfers and long-lived persistent sessions.
- DGA identification detects algorithmically generated domains using consonant-to-vowel ratios and length analysis.
- Log normalization processes Zeek (TSV/JSON), NetFlow, firewall, and proxy exports via CSV alias matching.
How it works
- Ingest telemetry by providing a
conn.log,dns.log, or a CSV export from your firewall or proxy. - Execute hunt scripts to score connection pairs based on interval and payload size consistency.
- Review findings sorted by threat score, focusing on high-regularity beacons and DGA-like domains.
- Map to ATT&CK to classify findings under techniques like T1071 (C2) or T1048 (Exfiltration).
Frameworks & tools
This tool uses Python for statistical analysis (RITA-style scoring). It supports Zeek (Bro) logs, NetFlow data, and generic CSV/TSV exports from enterprise firewalls.
Why this beats prompting it yourself
Writing prompts to find beacons often fails because LLMs struggle with the high-cardinality math required for jitter and coefficient of variation. This skill uses deterministic Python scripts to calculate regularity scores, ensuring you find actual periodic patterns rather than just guessing.
Use cases
- Threat hunting to find stealthy C2 channels in historic Zeek connection logs.
- Incident response to quickly identify the scope of data exfiltration after a breach.
- Compromise assessment to scan for DNS tunneling or DGA activity on a new network.
Known limitations
Large log files may require pre-filtering or chunking. Performance depends on the compute resources available to the agent environment.
Large datasets may require pre-filtering. Requires specific headers or Zeek formats for automatic column mapping. Results require manual verification against known update services.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago