v0.3.0 — 78 Skills, 23 Categories, Full Wireless Suite

What's New

Major expansion — claude-red goes from 37 skills in flat directories to 78 skills across 23 organized categories. This release adds full-spectrum offensive coverage from container escapes to wireless protocol attacks.

20 New Skills Across 10 Categories

• container/ — Docker/containerd breakout, runc CVEs, K8s RBAC abuse, etcd, kubelet API, pod escape
• • cicd/ — GitHub Actions injection, Jenkins RCE, GitLab CI, secret extraction, OIDC federation
• • api/ — OWASP API Top 10, REST/gRPC/WebSocket, batching attacks, webhook hijacking
• • crypto/ — Padding oracle, ECB, hash extension, RSA, POODLE/DROWN/Heartbleed, pinning bypass
• • privesc/ — Linux SUID/capabilities/sudo/kernel + Windows Potato family, DLL hijacking, UAC bypass
• • post-exploitation/ — Lateral movement (PTH/PTT/NTLM relay), persistence (Golden/Silver tickets), data exfiltration (DNS/HTTPS/ICMP tunneling)
• • forensics/ — Anti-forensics (timestomping, ADS, anti-VM) + C2 frameworks (Cobalt Strike, Sliver, Mythic, Havoc)
• • supply-chain/ — Dependency confusion, typosquatting, build pipeline attacks
• • social-engineering/ — GoPhish, EvilGinx2, MFA phishing, pretexting, vishing, USB drops
• • network/ — ARP spoofing, LLMNR/NBT-NS poisoning, VLAN hopping, IPv6 attacks, MITM

Wireless Suite — 14 Dedicated Skills

Split the monolithic offensive-wifi into per-surface skills covering the full wireless attack surface:

wifi-recon · wpa2-psk · wpa3-sae · wpa-enterprise · wps · evil-twin · krack-fragattacks · deauth-disassoc · bluetooth-ble · bluetooth-classic · zigbee-thread-matter · z-wave · lorawan-sub-ghz · wifi (comprehensive)

Deep Rewrites (2-4x More Content)

• offensive-deserialization — 183 → 781 lines. Java/PHP/.NET/Python/Node/Ruby, K8s vectors, WAF bypass
• • offensive-graphql — 207 → 946 lines. Introspection bypass, batching, subscription abuse, CrackQL/BatchQL
• • offensive-advanced-redteam — 146 → 640 lines. C2 tradecraft, OPSEC, evasion, engagement planning
• • offensive-ssti — 347 → 758 lines. 10+ engine-specific exploits, filter bypass, blind exploitation, chaining

Infrastructure & Tooling

• Skills reorganized into 23 category subdirectories
• • claude-skills.json manifest for programmatic access
• • install.sh one-liner installer
• • convert_skills.py for Claude Console compatibility (community contribution)
• • GitHub Actions: CodeQL scanning + automated skill review on PRs
• • SECURITY.md, CONTRIBUTING.md, LICENSE (MIT)

Community Contributions

Thanks to @yogesh-tessl for the skill-review GitHub Action and @LvL23HT for the bulk skill converter.

---

git clone https://github.com/SnailSploit/claude-red ~/.claude/skills/claude-red