Building AI-first systems, terminal-native skills, MCP servers, and operator tooling for Claude Code, Codex, and Gemini CLI.
[ TERMINAL INTAKE ]
I build fast, opinionated systems with an AI-first profile: skills, agent workflows, MCP infrastructure, and secure automation.
$ whoami
StealthyLabsHQ
$ active_profile
AI-first builder // skill engineer // MCP operator // security-minded
$ primary_agents
Claude Code / Codex / Gemini CLI
$ claude_code_status
100% operational mastery
$ codex_status
100% operational masteryThis profile is not meant to read like generic developer bio. It should feel like entry point into operating model:
- ship useful systems, not empty demos
- automate repeatable leverage
- secure early, not after launch
- make tooling feel sharp, deliberate, and alive
[ IDENTITY MATRIX ]
| FIELD | VALUE | STATUS |
|---|---|---|
| CODENAME | StealthyLabsHQ |
ACTIVE |
| UNIT | StealthyLabs |
ACTIVE |
| OPERATING MODE | AI systems + skills + infra + security | LIVE |
| CLAUDE CODE STATUS | 100% operational mastery | VERIFIED |
| CODEX STATUS | 100% operational mastery | VERIFIED |
| SIGNAL | fast builds, deep iteration, zero fluff | CONFIRMED |
[ AGENT STACK ]
| AGENT | OPERATIONAL ROLE | STATUS |
|---|---|---|
 Claude Code |
terminal-native skills, workflows, autonomous build loops | VERIFIED |
 Codex |
deep execution, debugging, shipping, security review | VERIFIED |
 Gemini CLI |
research acceleration, second-pass reasoning, contrast checks | ACTIVE |
[ FEATURED OPS ]
| REPO | BRIEF | STACK |
|---|---|---|
| caveman-lang | Token survival protocol — cut LLM output waste 60–90% | Prompt engineering |
| security-hardening | Portable defensive AppSec + AI-agent security skill (OWASP, MCP, prompt injection) | Python · Skill |
| gemini-mcp-connect | MCP bridge — Claude Code talks to Gemini brain | Python · MCP |
| steam-mcp | MCP server exposing Steam Web API to Claude Code & Gemini CLI | TypeScript · MCP |
| powerbi-mcp-local | MCP server for Power BI Desktop — Excel, Power Query, semantic model automation | Python · MCP |
| agentops-config-toolkit | Compact configs keeping Claude Code, Codex, Gemini focused + cost-efficient | Config · Skill |
[ NOW BUILDING ]
> expanding MCP server fleet — new integrations on the pipeline
> sharpening caveman-lang token rules across multi-agent stacks
> growing security-hardening skill for MCP + prompt injection defense
> full Discord RPC coverage for Claude Code + Codex operator sessions
> operator-grade skill templates for Claude Code / Codex / Gemini
[ VERIFIED CLEARANCES ]
[ TERMINAL DOCTRINE ]
> build systems, not noise
> forge skills for terminal-native agents
> prototype with intent
> automate leverage
> secure before scale
