auto_majicly

Self-taught dev & security researcher. Building autonomous pen testing agents on local LLMs. Founder @ Halo Cybersecurity. No cloud. No limits.

Arizona/United States Joined on Jun 29, 2026 christopherwsheridan@gmail.com https://github.com/XenoCoreGiger31/https://huggingface.co/automajicly/https://hashnode/HaloCybersecu

Education

Ashland University — Communications / Small Business

Pronouns

Work

Founder @ Halo Cybersecurity | Self-taught security researcher | Communications student | Building

Weak heuristics caused hallucinated success

auto_majicly

Jul 23

My AI pentest agent reported 23 root shells. It had actually popped zero.

#ai #cybersecurity #testing #mcp

8 min read

Want to connect with auto_majicly?

Create an account to connect with auto_majicly. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

auto_majicly

Jul 21

HALO v2.7: One Registry, 29 Tools, and a More Reliable MCP Architecture

#ai #python #mcp #security

2 min read

auto_majicly

Jul 18

From Tool-Runner to Decision Engine: Upgrading an Autonomous Security Agent's Harness

#ai #cybersecurity #python #opensource

5 min read

auto_majicly

Jul 13

Anatomy of a Pentest Dropbox: How Remote Internal Assessments Actually Work

#ai #pentest #cybersecurity #mcp

3 min read

auto_majicly

Jul 12

# Online vs. Offline Password Attacks: A Field Guide

#ai #hacktoberfest #cybersecurity #mcp

3 min read

auto_majicly

Jul 10

# Getting Listed: Publishing an Offensive-Security Agent to the MCP Registries

#ai #mcp #python #cybersecurity

3 min read

auto_majicly

Jul 9

I Built a Fully Local, Autonomous AI Pentesting Agent — Now I’m Teaching It to Speak MCP

#ai #cybersecurity #mcp #python

4 min read

auto_majicly

Jul 8

Your autonomous agent will lie to you about success

#discuss #cybersecurity #ai #llm

4 min read

auto_majicly

Jul 8

# Wiring Real Agents Into Halo: From Stub Orchestrator to a Working Multi-Agent Pentest Pipeline

#discuss #ai #programming #cybersecurity

2 min read

auto_majicly

Jul 7

SkillCloak Exposes a Dangerous Blind Spot in AI Security

#discuss #ai #security #hacktoberfest

1 min read

auto_majicly

Jul 7

Ai is doing it again: Strong armed by AI for Bitcoin

#discuss #cybersecurity #hacking #security

4 min read

auto_majicly

Jul 6

Building Halo’s Multi-Agent Architecture: From Single Loop to Five Specialists

#ai #security #agents #cybersecurity

2 min read

auto_majicly

Jul 6

🧠 Tokenization in Modern LLMs: The Hidden Mechanics Behind AI Costs and Reasoning

#discuss #ai #tokenizers #tutorial

4 min read

auto_majicly

Jul 6

How I stopped paying for AI

#showdev #ai #security #claude

1 min read

auto_majicly

Jul 5

The LLM Cost Death Spiral (And How I Got Out of It)

#showdev #opensource #python #security

5 min read

auto_majicly

Jul 5

Halo Is Going Multi-Agent: Splitting One Brain Into Five

#ai #security #cybersecurity #hacking

4 min read

auto_majicly

Jul 4

Someone Built a Physical Gear Shifter for Claude — and It's a Better UX Lesson Than Most Software Ships

#ai #cybersecurity #autonomy

3 min read

auto_majicly

Jul 4

Chasing the Light: How the June Solstice Game Jam Turned One Prompt Into a Hundred Different Games

#ai #gaming #solstice #developer

3 min read

auto_majicly

Jul 3

The Future of Agentic AI Memory Systems

#anonymous #ai #memory #cybersecurity

5 min read

auto_majicly

Jul 3

Prompt Injection Isn’t Going Away — Jason Haddix on the Architecture Problem Nobody Wants to Admit

#ai #interview #llm #security

3 min read

auto_majicly

Jul 2

Autonomous Pentesting Is Closer Than Expected — Jason Haddix on Where the Ceiling Actually Is

#pentesting #ai #cybersecurity #mcp

4 min read

auto_majicly

Jul 1

GEMMA-by-GOOGLE

#ai #cybersecurity #ethicalhacking #autonomous

2 min read

auto_majicly

Jul 1

Tool #23: Teaching Halo to Play Detective with Sherlock

#ai #cybersecurity #aitool #pentest

4 min read

auto_majicly

Jun 30

The Bug That Made My Pentesting Agent Give Up

#ai #autonomous #cybersecurity #pentest

3 min read

auto_majicly

Jun 29

GEMMA-by-GOOGLE

#ai #cybersecurity #opensource #python

2 min read

DEV Community

auto_majicly

Badges

2 Week Community Wellness Streak

1 Week Community Wellness Streak

GitHub Repositories

Skills/Languages

Currently learning

Currently hacking on

Available for

My AI pentest agent reported 23 root shells. It had actually popped zero.

Want to connect with auto_majicly?

HALO v2.7: One Registry, 29 Tools, and a More Reliable MCP Architecture

From Tool-Runner to Decision Engine: Upgrading an Autonomous Security Agent's Harness

Anatomy of a Pentest Dropbox: How Remote Internal Assessments Actually Work

# Online vs. Offline Password Attacks: A Field Guide

# Getting Listed: Publishing an Offensive-Security Agent to the MCP Registries

I Built a Fully Local, Autonomous AI Pentesting Agent — Now I’m Teaching It to Speak MCP

Your autonomous agent will lie to you about success

# Wiring Real Agents Into Halo: From Stub Orchestrator to a Working Multi-Agent Pentest Pipeline

SkillCloak Exposes a Dangerous Blind Spot in AI Security

Ai is doing it again: Strong armed by AI for Bitcoin

Building Halo’s Multi-Agent Architecture: From Single Loop to Five Specialists

🧠 Tokenization in Modern LLMs: The Hidden Mechanics Behind AI Costs and Reasoning

How I stopped paying for AI

The LLM Cost Death Spiral (And How I Got Out of It)

Halo Is Going Multi-Agent: Splitting One Brain Into Five

Someone Built a Physical Gear Shifter for Claude — and It's a Better UX Lesson Than Most Software Ships

Chasing the Light: How the June Solstice Game Jam Turned One Prompt Into a Hundred Different Games

The Future of Agentic AI Memory Systems

Prompt Injection Isn’t Going Away — Jason Haddix on the Architecture Problem Nobody Wants to Admit

Autonomous Pentesting Is Closer Than Expected — Jason Haddix on Where the Ceiling Actually Is

GEMMA-by-GOOGLE

Tool #23: Teaching Halo to Play Detective with Sherlock

The Bug That Made My Pentesting Agent Give Up

GEMMA-by-GOOGLE