Multi-Platform Agentic AI Environment & Skill Suite

Production-ready, cross-platform configuration framework, system rules (Prime Directives), 24 modular skills (agentskills.io standard), and MCP configurations for Antigravity / Gemini CLI, Claude Code, and Modern Cursor IDE (.mdc).

---

Overview

Modern AI coding agents often suffer from common failure modes:

1. Simulation & Mocking Traps: Writing fake stubs or untested boilerplate without running tests.
2. Context Window Flooding: Inefficiently dumping thousands of prompt lines instead of on-demand skill discovery.
3. Monolithic Spaghetti Code: Creating giant files without clean architecture or boundary isolation.
4. Hardware & Version Hallucinations: Speculating about hardware capabilities or package versions without live verification.

agent-setup-bundle solves this by establishing a unified, multi-platform engineering operating system across your favorite AI coding assistants.

---

Key Features

1. 10 Operational Core Rules (~/.agents/rules/)

• modular-architecture.md — Anti-Monolith constraint (max ~150-200 lines), Clean/Hexagonal architecture, strict layer direction, typed boundary contracts.
• zero-speculation.md — Total ban on guessing versions, APIs, or system states without live verification.
• systemic-excellence.md — Anti-workaround protocol, single source of truth (SSOT), and DRY enforcement.
• system-identity.md — Hardware & OS baseline template to prevent environment hallucinations.
• subagent-economy.md — Dynamic tiering of LLM models based on task complexity.
• problem-isolation.md — Strict scope containment and root cause debugging.
• command-verification.md — Mandatory read-only verification after running commands.
• env-integrity.md — Environment sanity checks before modifying project files.
• error-triage.md — Diagnostic priority order (Docs -> Web -> Code).
• full-log-reporting.md — Ban on truncated or paraphrased error reporting.

2. 24 Production-Grade Skills (~/.agents/skills/)

All skills adhere to the open agentskills.io standard with progressive disclosure:

Skill	Description & Trigger
skill-monorepo-architect	Polyglot monorepo management (uv Python workspaces + pnpm/Turborepo), shared-core pattern.
skill-plugin-architecture	Microkernel architecture, BasePlugin, dynamic discovery, and fault isolation.
skill-osint-engineering	Threat intelligence gathering, Pydantic entity graphs, pivoting, and OPSEC protocols.
skill-stealth-scraping	Anti-bot evasion (Cloudflare/DataDome), TLS fingerprint spoofing, headless browser automation.
spec-driven-development	Gated SDLC (Specify -> Plan -> Tasks -> Implement).
skill-backend-architect	Contract-first API design, database schemas, Pydantic/Zod boundaries.
skill-frontend-architect	Clean UI/UX architecture, state separation, accessibility, and performance.
skill-qa-engineer	Test-Driven Development (TDD Red-Green), TypeScript compilation safety gates.
doubt-driven-development	Fresh-context adversarial reviews to challenge false confidence.
skill-codebase-onboarding	Cartography and progressive disclosure for unfamiliar repositories.
skill-code-review	5-axis systematic code review (correctness, readability, architecture, security, performance).
skill-system-diagnostics	Kernel, OS, hardware, and driver diagnostics.
skill-devops-cloud	Docker containers, CI/CD pipelines, and cloud infrastructure.
skill-data-science	Data ingestion pipelines, exploratory data analysis, and validation.
skill-data-analysis	Statistical analysis, hypothesis testing, and anomaly detection.
skill-ai-ml	LLM model integrations, embeddings, vector databases, and agent workflows.
skill-graph-analytics	Graph databases (Neo4j), topology analysis, and Graph Data Science.
skill-graphics-webgl	2D/3D WebGL scenes, Three.js, and Canvas animations.
marketing-copywriting	Conversion-focused technical copywriting.
avoid-ai-writing	Detection and removal of AI writing clichés and robotic patterns.
seo-optimization-and-audit	Web metadata, Core Web Vitals, and search ranking optimization.
skill-research	Rigorous technical and academic multi-source research.
skill-resume-tailor	ATS-optimized resume, CV, and cover letter tailoring.
skill-web-architecture	Full-stack web architectural standards and API contracts.

---

1-Click Quickstart Installation

Clone and run the installer in your environment:

git clone https://github.com/Gzyms69/agent-setup-bundle.git
cd agent-setup-bundle
./install.sh

Installation Options:

./install.sh --all       # Configure all platforms (default)
./install.sh --gemini    # Configure Antigravity / Gemini CLI only
./install.sh --claude    # Configure Claude Code only
./install.sh --cursor    # Configure Cursor IDE only

---

Configuration & Setup

1. Configure API Keys:
   • Update ~/.gemini/settings.json or ~/.cursor/mcp.json with your GitHub token (GITHUB_PERSONAL_ACCESS_TOKEN) or Google API Key if using MCP servers.
2. System Baseline:
   • Customize ~/.agents/rules/system-identity.md with your system's hardware specs (CPU, GPU, RAM) to prevent the AI from guessing system capabilities.

---

Supported Environments

Platform	Manifest Location	Rule Format	Skill Support
Antigravity / Gemini CLI	~/.gemini/GEMINI.md	~/.agents/rules/*.md	~/.agents/skills/
Claude Code	~/.claude/CLAUDE.md	CLAUDE.md + .claude/agents/	~/.agents/skills/
Cursor IDE	Root / Project	.cursor/rules/*.mdc	On-demand docs

---

License

This project is licensed under the MIT License.