francescostabile/numasec/forensics-kit
- community
Checking that the SKILL.md file is still in the repository.
Digital forensics and incident response toolbox. Load when the operator asks about a pcap, a binary, a memory dump, a suspicious file, malware triage, IOC hunting, or post-incident analysis. Covers network (tshark), binaries (radare2, strings, binwalk, file, exiftool), memory (volatility), and pattern matching (YARA). All commands assume the artifact is local and disposable; never analyze in-place on a production system.
Install
npx vskill@latest install francescostabile/numasec/forensics-kitSKILL.md
Forensics & IR kit
Rule of engagement. Forensics work runs on copies. Never analyze a live victim system in place — cp/dd to a workspace, hash the original, and work on the copy. Every finding must reference the SHA256 of the artifact so chain-of-custody survives.
Output contract. Each IOC, carved artifact, or suspicious observation is a candidate observation (severity medium+ for IOCs, confidence high for exact matches, medium for heuristic hits).
When to load
| Situation | Why | |
|---|---|---|
Operator drops a .pcap, .pcapng, .bin, .dmp, .elf, .exe | You need these tools | |
| Post-compromise triage of a Linux/Windows host image | Memory + binaries | |
| Suspicious file discovered by a scanner | strings / binwalk / file first | |
| IOC hunt across a dataset | YARA rules |
Toolbox
1. file + exiftool + strings — first-3-minutes triage
Before anything heavier, always identify the artifact and skim it.
Interpretation. file gives format/arch. exiftool surfaces author/tool metadata. strings reveals hard-coded URLs, C2 hostnames, file paths, error messages, library imports.
2. binwalk — Carve embedded content
Severity hints. Firmware that carves cleanly to a shell script + a busybox → high if it contains credentials. Very-high uniform entropy → encrypted blob; flag and move on.
3. tshark — Packet capture analysis
Interpretation. Unknown SNI + sustained beacon interval ≈ C2 candidate. Non-RFC1918 destinations in an internal capture → exfiltration candidate.
4. volatility (Volatility 3) — Memory forensics
Severity hints. malfind hit with RWX region and MZ header → high, confidence medium. hashdump success → critical (credential exposure).
5. radare2 — Binary RE
For one-liners:
6. YARA — Signature-based hunting
Good rulesets to start from (no key required):
- YARA Rules project: https://github.com/Yara-Rules/rules
- Neo23x0 Signature-Base: https://github.com/Neo23x0/signature-base
7. Hashing + Chain-of-custody hygiene
Also keep a copy of ls -la artifact.bin and stat artifact.bin so mtime/ctime are recorded.
Standard triage opening
For an unknown artifact:
sha256sum+file+exiftool+strings | head.- If it's a binary:
binwalk -E(entropy) →binwalk -eif entropy is moderate →radare2 -Aif interesting. - If it's a pcap:
tshark -z io,phs→ HTTP/DNS dumps → SNI → export objects. - If it's a memory image:
vol windows.info→pslist→netscan→malfind→hashdump. - Run YARA with a generic ruleset.
- Summarize IOCs (hosts, hashes, paths, mutexes, registry keys) into observations.
- Cross-reference IOCs against
cvetool (T10) and any open-source threat intel the operator has.
What this skill does NOT do
- No live acquisition. Operator pre-acquires memory/disk.
- No cloud forensics (AWS CloudTrail parsing, GCP audit logs). Those need a dedicated skill pack.
- No automated malware detonation / sandboxing. Use Cuckoo/CAPE externally if needed.
Escalation
After triage:
- If a binary has exploitable defects → hand off to the
hackingagent viatasktool. - If an IR finding needs remediation code changes → hand off to the
appsecagent. - If the pcap shows sustained C2 → open a parallel threat-intel subtask (
securityagent).
Security
- Security scanPassed, 91 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 repositoryCheckingpackages/numasec/skills/forensics-kit/SKILL.md
Popularity
- Repository stars
- 372
- Repository forks
- 47
- 7-day trend score
- 30
Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 24 May 2026, 5 months ago.
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/francescostabile/numasec/forensics-kit)