Agentic AI Solution — AI skill for Claude Code
Hands-on Agentic AI solution: a simple local multi-model agent with MCP tools and long-term memory that is capable of reading and writing files inside its sandbox space using local LLM like Gemma4 mod.
How to install Agentic AI Solution
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open MohamedArzath/agentic-ai-solution and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Agentic AI Solution does
Hands-on Agentic AI solution: a simple local multi-model agent with MCP tools and long-term memory that is capable of reading and writing files inside its sandbox space using local LLM like Gemma4 model and cloud models like Claude and ChatGPT.
Alternatives in AI
- OpenSpace — "OpenSpace: The Skill Management Layer for AI Agents" -- https://open-space.cloud/ 7.5k ★
- Hindsight — State-of-the-art long-term memory for AI agents by Vectorize 6.7k ★
- J Space Cognition Suite V3.7 — J-Space Cognition Suite V3.7 - AI cognitive-enhancement Skills based on Anthropic's J-space global workspace r 3k ★
README
🤖 Agentic AI Solution
**A from-scratch AI agent harness with local and cloud models, file tools, MCP servers and long-term memory**
[](https://python.org) [](https://docs.astral.sh/uv/) [](https://ollama.com) [](https://modelcontextprotocol.io) [](https://platform.openai.com/docs/api-reference/chat) [](LICENSE)
A small, readable AI agent harness built from scratch on the OpenAI-compatible Chat Completions API. It contains two agents:
| Script | What it is |
|---|---|
| `main.py` | Simple agent. One local model (Ollama), a tool-calling loop and three file tools scoped to src/workspace/. |
| `agentic_ai.py` | Full agent. Several models you can switch at runtime (local and cloud), MCP server tools, long-term memory and slash commands. |
Both agents use the same plain loop: send the conversation to the model, run any tools it asks for, append the results and repeat until the model gives a final answer.
Features
`main.py`: simple local agent
- Talks to a local Ollama model (
gemma4:12bby default) through its OpenAI-compatible endpoint - File tools:
list_files,get_file_content,create_file - All file access stays inside
src/workspace/ - Each prompt starts a new conversation, with no state kept between turns
`agentic_ai.py`: multi-model agent
- Model registry. Local Ollama models (Gemma, Qwen, Phi) plus
Related Skills
Planner System
You are Axion, a general-purpose AI assistant running on macOS. You are equally capable at software engineerin
Orvima
Why should Claude be the only one with a browser? Orvima gives any MCP-capable AI (Claude, Cursor, Copilot, yo
AI Agent Sandbox
A secure bubblewrap-based sandboxing solution for running Claude Code with strict filesystem isolation.
Claude Code Redact
Claude Code mod that redacts secrets and PII before the model reads them, stores placeholders in the conversat
Agentacl
Access control and sandboxing for AI coding agents on macOS. Stop Claude Code, Codex, Gemini CLI, Copilot CLI
Nocturne Memory
A lightweight, rollbackable, and visual Long-Term Memory Server for MCP Agents. Say goodbye to Vector RAG and
Related Agents
Memory Searcher
Searches the long-term memory vault and returns a dense answer with path:line citations. Use for broad questio
Context Memory Manager
Context Window & Token Optimization Specialist. Manages token budgets, compresses prompts (RTK/Caveman), imple
Probe
Deep-inspects a media file and reports what it actually is, not what the filename claims. Use when the user as
