malware-triage — Claude Code skill: analyze and classify malware
malware-triage — a skill for safe analysis of unknown malware samples. Extracts static indicators, runs code in a sandbox, identifies behavior patterns, and produces detection guidance with IOCs and ATT&CK mapping.
no license 234 installs updated 2026-08-26
npx skills add securityskills/skills --skill malware-triage -a claude-codeWhat malware-triage does
The skill automates malware analysis from start to finish. It begins with static examination: compute hashes, identify file type, scan strings for URLs and C2 patterns, validate digital signatures, detect packers. Next, detonate the sample in an isolated VM or sandbox while monitoring file system, registry, network, and process activity—capture created files and any command-and-control communication.
Then extract behavior patterns: what persistence mechanisms are set up, which injection techniques are used, how data flows out, what defenses are bypassed. Output includes family attribution through YARA rules, a mapping to ATT&CK techniques, machine-readable IOCs, and a one-page summary that tells you whether deeper reverse engineering is needed or if behavioral signals are enough.
Requires an isolated sandbox (VM, Cuckoo, ANY.RUN, or vendor platform) and familiarity with analysis scripts. Samples must be stored in password-protected archives and never accessed outside the sandbox.
What it helps with
- classify malware by family and behavior pattern
- extract IOCs and indicators of compromise in hours instead of days
- identify persistence, injection, and data exfiltration methods
- generate YARA rules and detection recommendations
- decide whether deep disassembly is worth the time or behavioral analysis suffices
Example prompts for the agent
- Analyze the sandbox report, extract IOCs and map behavior to ATT&CK techniques
- Triage a file — static analysis, dynamic execution, persistence checks, write summary
- Build a YARA rule for the malware family based on strings and hashes
How to install malware-triage
1Skills CLI (Vercel)
npx skills add securityskills/skills --skill malware-triage -a claude-code2Manually: put the files in place
~/.claude/skills/malware-triage/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/malware-triage/
1Skills CLI (Vercel)
npx skills add securityskills/skills --skill malware-triage -a codex2Manually: put the files in place
.agents/skills/malware-triage/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add securityskills/skills --skill malware-triage -a cursor2Manually: put the files in place
.cursor/skills/malware-triage/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install securityskills/skills malware-analysis/malware-triage/SKILL.md2Manually: put the files in place
.github/skills/malware-triage/SKILL.md1Skills CLI (Vercel)
npx skills add securityskills/skills --skill malware-triage -a gemini-cli2Platform CLI
gemini skills install https://github.com/securityskills/skills --path malware-analysis/malware-triageManually: put the files in place
.devin/skills/malware-triage/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add securityskills/skills --skill malware-triage -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about malware-triage
How do I install malware-triage in Claude Code?
Run npx skills add securityskills/skills --skill malware-triage -a claude-code; the skill will load into ~/.claude/skills/malware-triage/SKILL.md and become available in Claude Code.
What does malware-triage do?
It guides you through safe dissection of an unknown file: scan for malicious indicators, execute it in isolation, extract behavior (persistence, code injection, exfiltration), and generate a report with IOCs, YARA rules, and next steps.
What sandbox environment do I need?
An isolated VM or dedicated sandbox (Cuckoo, ANY.RUN, or similar) with network simulation or sinkholing. Store samples in password-protected archives and run only from snapshots that you revert after each execution.