shady @_shday

AI-Powered Red Team — 28 Specialized Agents for Offensive Security 🤖🔥 Turn Claude into a full pentesting team. • 28 agents (Recon, AD, Web, Cloud, Mobile) • Auto task routing → correct agent • Real tools support (nmap, sqlmap, nuclei, BloodHound) • Recon → Exploit → Report 🔗 github.com/0xSteph/pentes… #artificialintelligence #RedTeam #Pentesting #cybersecurity #infosec

754 structured cybersecurity skills mapped to MITRE frameworks for AI agents. A powerful resource for anyone working in cybersecurity or building AI agents for offense, defense, and protection. Check out the full repo here: github.com/mukul975/Anthr… ⭐️

⚔️ Claude-Red = Offensive Security Skills for Claude AI A massive open-source framework that transforms Claude into a context-aware red team assistant. 🔥 📚 100+ offensive security skill modules 🌐 Web exploitation 🧠 Active Directory attacks ☁️ Cloud attack paths 📡 Wireless exploitation 💥 Exploit dev & fuzzing 🤖 AI security testing Built for: 🎯 Bug bounty hunters 🛡️ Red teamers 🔬 Security researchers 🎓 CTF players ☁️ Cloud pentesters Capabilities include: ✅ SQLi / XSS / SSRF / RCE ✅ ADCS / Kerberos / ACL abuse ✅ EDR bypass & shellcode ops ✅ WPA2/WPA3 attacks ✅ OAuth & JWT exploitation ✅ Prompt injection & jailbreak testing 🔗 github.com/SnailSploit/Cl… #RedTeaming #Hacking #CyberThreat #ThreatIntel #Pentest #OpenSource

Claude Code Skill Bundle for Bug Bounty Hunting & External Red Team Operations 🤖💀 • 51 offensive security skills + 15 slash commands • Trained on 574+ disclosed HackerOne-style report patterns • Covers XSS, SSRF, SQLi, OAuth, JWT, GraphQL, RCE, IDOR & API abuse • Enterprise attack chains for M365, Okta, VPNs, SharePoint & vCenter • Built-in recon, exploit chaining, triage, evidence hygiene & reporting github.com/elementalsouls… #BugBounty #RedTeam #CyberSecurity #Pentesting #AppSec

4 levels of Hermes Agent setup: LEVEL 1: main agent You → Hermes Agent this is your main agent and your prototype area, where you test new workflows and refine them. it doubles as your orchestrator until you have something worth breaking out ---- LEVEL 2: specialized agents You → SEO Agent You → CMO Agent You → Ops Agent once a workflow is solid, break it out into its own agent with its own credentials, memory and scope. --- LEVEL 3: orchestrated team You → Orchestrator ↓ Specialist Agents bring the orchestrator back in. it now steers the company of agents you have built. ---- LEVEL 4: automated team Cron / Events → Orchestrator ↓ Agent Team add task lists so the team works async. cron and events fire jobs, the orchestrator routes them through the task bus, the team handles the work without you ---- take small steps, you DO NOT want to automate slop. if your output at level 1 is mediocre, you are about to scale mediocrity. 20 agents shipping low quality work at speed is worse than 3 shipping great work slowly. I would rather run fewer agents with better output than MAXXING the agent count and spitting out more of the same.

Shann³ @shannholmberg

3 weeks ago

x.com/i/article/2055…

35 245 2K 1.4M 7K

I've spent the past few weeks rebuilding my entire marketing operation in Hermes Agents I poured everything I learned along the way into this article. > how to set it up > the four-level path from one agent to a whole marketing company on one VPS > the agent control room template I built (public repo to clone) > the models I run for creative vs structured work go become an operator

Shann³ @shannholmberg

3 weeks ago

x.com/i/article/2055…

35 245 2K 1.4M 7K

Give this a try for subdomain takeover #bugbountytips #bugbounty github.com/mikaww1/Subdom…

🚨 JUST IN - Google published a long piece about "Optimizing your website for generative AI features on Google Search" 👀 A lot in it developers.google.com/search/docs/fu… 🧵

Nice list thanks for sharing. I’d add 3 “operating system” workflows if I were also starting from scratch: Strategic Memory Curator
Decides what Hermes should remember, archive, expire, or ignore. Prevents the second brain from becoming a second landfill. Promise Tracker
Captures every “I’ll send,” “I’ll follow up,” “I’ll check,” “I owe you,” etc. Turns trust leaks into an execution queue. Weekly Kill List
Every Friday, Hermes recommends what to kill, pause, delegate, simplify, merge, or automate. Keeps the system from accumulating bloat. Together: Memory Curator = keeps the brain clean
Promise Tracker = keeps trust intact
Kill List = keeps the system lean

If I was starting Hermes from zero, these are the 9 workflows I'd build first (to make it a real Chief of Staff): 1. Daily Brief Every morning at 7am, Hermes pulls my calendar, top 3-5 urgent emails, weather, and 3 headlines from my interest feeds, then drops it as one scannable message in Telegram. Replaces my old shitty ritual of opening 5 apps before coffee. 2. Viral Swipe File (self-improving) A nightly cron checks every post I've published across X, LinkedIn, and Threads. Anything that crosses my engagement threshold gets auto-extracted into a structured swipe file with the hook, structure, topic, opening line, and stats. It gets better every week. Over time the swipe file builds a precise fingerprint of what works for me, calibrated against real data. 3. Trending Workflows Radar Every morning Hermes scans Reddit, X, and AI forums, identifies what workflows are gaining velocity in the last 24 hours, and delivers a ranked list of 5 content angles. This helps me stay on top of the hottest workflows people are cooking in AI. 4. Meeting Prep Briefing 30 minutes before every Google Calendar meeting, Hermes pulls the attendee list, fetches their LinkedIn/company context, summarizes my last email thread with them, and sends a one-page brief to Telegram. I walk into every call sounding prepared without digging through threads. 5. The Humanizer A skill that audits any text against 30+ known AI writing tells (em-dashes, "delve," "tapestry," tricolon structures) and rewrites them into natural prose. Lets me accelerate first drafts with AI without sounding like I did (probably my most used workflow in my entire stack) 6. Bookmark Inbox Hermes monitors my X bookmarks automatically. Anything new gets fetched, summarized in 3 bullets, auto-tagged, and filed into my Obsidian vault by topic. Saved stuff becomes searchable knowledge instead of digital clutter. 7. Customer Support Cron Every morning Hermes scans my inbox for support tickets, categorizes them by issue type, and logs everything to my company Discord. Weekly report surfaces the top 5 recurring issues so I know what to actually fix in the product. 8. Weekly Business Report Every Monday morning Hermes pulls Stripe revenue, newsletter subs, content views, follower growth, churn, and refunds. Then drops it as a single dashboard in Telegram with this-week-vs-last-week. 9. Obsidian LLM Wiki Second Brain A single Obsidian vault that becomes the source of truth for everything in my business / life (Karpathy-maxxing) I have Hermes writes a daily report on everything that happened across my Discord and Telegram, then add it to the vault. Over time it becomes a deep knowledge base I can point any model at. ••• If you want to build these, simply paste this post into your Hermes agent and tell it to build the ones you want. It'll ask you which integrations to connect (Gmail, Stripe, Telegram, etc), pull your business context, and set them up for you. What workflows do you love that should I add??

the anatomy of ~/.hermes folder. one folder controls everything your hermes agent knows, remembers, and can do. understanding its layout is the difference between treating hermes as a black box and actually customizing it. here's what lives inside and why each piece matters. 𝗰𝗼𝗻𝗳𝗶𝗴𝘂𝗿𝗮𝘁𝗶𝗼𝗻 𝗰𝗼𝗻𝗳𝗶𝗴.𝘆𝗮𝗺𝗹 is the source of truth for everything non-secret: model choice, terminal backend, tool enablement, MCP servers. 𝗲𝗻𝘃 holds your API keys and bot tokens. 𝗮𝘂𝘁𝗵.𝗷𝘀𝗼𝗻 stores OAuth credentials. then there's 𝗦𝗢𝗨𝗟.𝗺𝗱. it occupies slot #1 in the system prompt, before anything else loads. it defines who the agent is: personality, tone, communication style, hard limits. everything the agent writes, creates, and remembers passes through this identity layer. 𝗸𝗻𝗼𝘄𝗹𝗲𝗱𝗴𝗲 𝗺𝗲𝗺𝗼𝗿𝗶𝗲𝘀/ contains two tiny files. 𝗠𝗘𝗠𝗢𝗥𝗬.𝗺𝗱 (2,200 chars) holds project conventions, tool quirks, lessons learned. 𝗨𝗦𝗘𝗥.𝗺𝗱 (1,375 chars) holds your profile. both get injected into the system prompt as frozen snapshots at session start. when they fill up, the agent consolidates: merges entries, drops redundancy, keeps only what's dense and useful. 𝗰𝗮𝗽𝗮𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀 𝘀𝗸𝗶𝗹𝗹𝘀/ is where the learning loop lives. each skill is a self-contained ability: a 𝗦𝗞𝗜𝗟𝗟.𝗺𝗱 (the procedure), a 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲𝘀/ folder (docs the agent reads), and 𝘀𝗰𝗿𝗶𝗽𝘁𝘀/ (executable helpers). skills come from three sources: bundled with hermes, downloaded from the hub via 𝗵𝘂𝗯/, or created by the agent itself during your sessions. hermes ships with 687 skills across 18 categories, and you can add any GitHub repo as a custom tap. 𝗿𝘂𝗻𝘁𝗶𝗺𝗲 𝘀𝘁𝗮𝘁𝗲 𝘀𝗲𝘀𝘀𝗶𝗼𝗻𝘀/ stores per-platform session metadata. 𝘀𝘁𝗮𝘁𝗲.𝗱𝗯 is the SQLite database with FTS5 indexing that backs tier 2 memory. this is what makes "what did we discuss three weeks ago?" actually work across CLI and messaging. 𝗮𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻 𝗰𝗿𝗼𝗻/ holds scheduled jobs in 𝗷𝗼𝗯𝘀.𝗷𝘀𝗼𝗻 and their outputs in 𝗼𝘂𝘁𝗽𝘂𝘁/. the gateway daemon ticks every 60 seconds and runs due jobs in isolated sessions. you describe schedules in plain English, hermes converts them. 𝗲𝘅𝘁𝗲𝗻𝘀𝗶𝗼𝗻 + 𝗼𝗯𝘀𝗲𝗿𝘃𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗽𝗹𝘂𝗴𝗶𝗻𝘀/, 𝗵𝗼𝗼𝗸𝘀/, and 𝘀𝗸𝗶𝗻𝘀/ are the surface area for user customization. 𝗹𝗼𝗴𝘀/ gives you 𝗮𝗴𝗲𝗻𝘁.𝗹𝗼𝗴, 𝗴𝗮𝘁𝗲𝘄𝗮𝘆.𝗹𝗼𝗴, and 𝗲𝗿𝗿𝗼𝗿𝘀.𝗹𝗼𝗴 for debugging. you won't manually edit most of these files. but knowing this layout means you understand exactly where identity, memory, skills, automation, and state live, and how they connect. i wrote a full deep dive covering hermes agent's architecture, memory system, self-evolving skills, GEPA optimization, and setting up multiple specialized agents. The article is quoted below.

Akshay 🚀 @akshay_pachaar

3 weeks ago

x.com/i/article/2053…

113 751 6K 5.1M 22K

OSINT Drop: WhatsApp Intelligence (Legal & Ethical) A single phone number can unlock more than most people realize. → Profile photos (even historical changes) → Status messages & behavioral patterns → Last seen / activity windows → Linked metadata across platforms No exploits. No hacking. Just pure OSINT tradecraft. The real edge isn’t the tool — it’s the methodology: • Correlate number → identity • Pivot into social platforms • Track reuse patterns • Build behavioral timelines Most people leave a data trail without knowing it. If you know how to follow it… you don’t need access. Full guide 👇 hackers-arise.com/open-source-in…

Here's a cool trick for y'all looking to create new Nuclei templates for exploitable CVEs! Using CVEmap you can get a list of CVEs with public proofs of concept, that have been marked as exploitable by CISA, are remotely exploitable AND don't have a Nuclei template (yet)! Flags: -k / -kev: Marked as exploitable vulnerabilities by CISA -t=false / -template=false: Has no public Nuclei templates -poc: Has public published POC -re / -remote: is remotely exploitable Good luck! 🤞 #nuclei #hacking #pentesting #bugbounty #CVEmap

Imagine controlling a SWARM of HermesAgents. One control panel for: > Orchestration loops > Missions > Checkpoints > Review gates > Autopilot Run a team of 5, 10, or 100 agents 😂

Autonomous bug-bounty framework for Claude Code — 40 specialist agents, exploit-chain builder, writeup search, and live HackerOne/Bugcrowd integration. github.com/H-mmer/pentest…

CPU vs GPU vs TPU vs NPU vs LPU, explained visually: 5 hardware architectures power AI today. Each one makes a fundamentally different tradeoff between flexibility, parallelism, and memory access. > CPU It is built for general-purpose computing. A few powerful cores handle complex logic, branching, and system-level tasks. It has deep cache hierarchies and off-chip main memory (DRAM). It's great for operating systems, databases, and decision-heavy code, but not that great for repetitive math like matrix multiplications. > GPU Instead of a few powerful cores, GPUs spread work across thousands of smaller cores that all execute the same instruction on different data. This is why GPUs dominate AI training. The parallelism maps directly to the kind of math neural networks need. > TPU They go one step further with specialization. The core compute unit is a grid of multiply-accumulate (MAC) units where data flows through in a wave pattern. Weights enter from one side, activations from the other, and partial results propagate without going back to memory each time. The entire execution is compiler-controlled, not hardware-scheduled. Google designed TPUs specifically for neural network workloads. > NPU This is an edge-optimized variant. The architecture is built around a Neural Compute Engine packed with MAC arrays and on-chip SRAM, but instead of high-bandwidth memory (HBM), NPUs use low-power system memory. The design goal is to run inference at single-digit watt power budgets, like smartphones, wearables, and IoT devices. Apple Neural Engine and Intel's NPU follow this pattern. > LPU (Language Processing Unit) This is the newest entrant, by Groq. The architecture removes off-chip memory from the critical path entirely. All weight storage lives in on-chip SRAM. Execution is fully deterministic and compiler-scheduled, which means zero cache misses and zero runtime scheduling overhead. The tradeoff is that it provides limited memory per chip, which means you need hundreds of chips linked together to serve a single large model. But the latency advantage is real. AI compute has evolved from general-purpose flexibility (CPU) to extreme specialization (LPU). Each step trades some level of generality for efficiency. The visual below maps the internal architecture of all five side by side. 👉 Over to you: Which of these 5 have you actually worked with or deployed on?

Agent memory is three-dimensional. Most agent memory systems use a single store. Usually a vector database. It handles semantic similarity well, but it captures only one dimension of knowledge. Here's the gap. Store these three facts: → Alice is the tech lead on Project Atlas → Project Atlas uses PostgreSQL for its primary datastore → The PostgreSQL cluster went down on Tuesday Now ask: was Alice's project affected by Tuesday's outage? Vector search finds fact 1 (mentions Alice) and fact 3 (mentions Tuesday). But the bridge between them, fact 2, mentions neither. It connects Project Atlas to PostgreSQL, and that's exactly what gets missed. This is the normal shape of business knowledge. People belong to teams, teams own projects, projects depend on systems, systems have incidents. Any question crossing two hops breaks flat retrieval. The three dimensions that actually cover agent memory: → A relational store for provenance (where data came from, when, who has access) → A vector store for semantics (what content means, what it's similar to) → A graph store for relationships (how entities connect across hops) Each captures something the other two can't. Vectors find meaning. Graphs trace connections. Relational tables track lineage and permissions. The real unlock is combining them: enter through vectors (find semantically relevant content), then traverse the graph (follow edges to connected entities), with provenance grounding every result back to its source. Cognee is an open-source project that unifies all three behind four async calls. The default stack is fully embedded (SQLite + LanceDB + Kuzu), so a pip install gets you running locally. For production, swap in Postgres, Qdrant, or Neo4j without changing your agent code. Check it out on GitHub: github.com/topoteretes/co… The article below is a first-principle deep dive on building agents that never forget. This will give you a clear picture of how memory for agents is evolving.

Akshay 🚀 @akshay_pachaar

2 months ago

x.com/i/article/2043…

46 305 2K 624K 4K

My RAG Techniques book just became an Amazon #1 Bestseller in 24 hours! 🏆 📊 The Stats: • 630+ copies sold in ONE day. • #1 Bestseller in 3 categories. • Ranked #813 out of 12M+ Kindle books. To celebrate, I’m extending the $0.99 launch price (literally free) for just 24 more hours! 22 chapters teaching you everything about RAG, based on my 27k-star GitHub repo. Want the link? Reply "RAG" ♻️ Repost this so others can grab it! I'll DM the link to everyone who interacts. Thanks for the support!

🟥𝗟𝗟𝗠 𝗥𝗲𝗱 𝗧𝗲𝗮𝗺𝗶𝗻𝗴 𝗥𝗲𝘀𝗼𝘂𝗿𝗰𝗲𝘀 🔗 Red Teaming LLM Applications by DeepLearning :- deeplearning.ai/short-courses/… 🔗Planning red teaming for LLMs by Microsoft :- learn.microsoft.com/en-us/azure/ai… 🔗Red teaming LLMs by HuggingFace :- huggingface.co/blog/red-teami… 🔗LLM Red Teaming by KLU :- klu.ai/glossary/red-t… 🔗Red Teaming LLMs Nihad Hassan :- techtarget.com/searchenterpri… 🔗Red-Teaming to make LLMs robust and safer :- medium.com/@sasirekharame… 🔖#infosec #cybersecurity #hacking #pentesting #security

@NirDiamantAI And for the .com site, the ebook version is not available