Downloads · 30 days
11
30% of all-time downloads
beatsprom/qwen-stateless-mcp-agent-2026
qwen-stateless-mcp-agent-2026 is a text generation model from beatsprom. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as apache-2.0.
[](https://opensource.org/licenses/Apache-2.0) [](https://huggingface.co/datasets/beatsprom/stateless-mcp-agent-evaluation-suite-2026) [](https://beatsprom.gumroad.com/l/stateless-mcp-agent-eval)
Downloads · 30 days
11
30% of all-time downloads
All-time downloads
37
Public
Repo size
—
Likes
0
Public
Click a slice to open those files.
.md3.8 KB · 57%
From the Hugging Face model README
This model repository contains the official LoRA adapter and Ollama Modelfile fine-tuned on the Enterprise Stateless MCP & Agent Evaluation Suite (2026). It specializes in zero-defect, deterministic execution of Model Context Protocol (MCP 2026 JSON-RPC 2.0) tool calls across 25 production infrastructure domains.
| Benchmark Metric | Open Weights Baseline | Fine-Tuned (This Agent) | Absolute Delta |
|---|---|---|---|
| BFCL Tool-Calling Accuracy | 68.4% | 94.8% | +26.4% |
| SQL Injection & Schema Tampering Defense | 71.2% | 99.8% | +28.6% |
| Idempotency Key Strict Adherence | 62.0% | 99.9% | +37.9% |
| Rate Limit (HTTP 429) Circuit Recovery | 58.5% | 93.2% | +34.7% |
Run this agent locally on any Mac/Linux/Windows machine with 1 click:
# 1. Clone repository
git clone https://huggingface.co/beatsprom/qwen-stateless-mcp-agent-2026
cd qwen-stateless-mcp-agent-2026
# 2. Build and run in Ollama
ollama create stateless-mcp-agent -f Modelfile
ollama run stateless-mcp-agent
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
base_model_id = "Qwen/Qwen2.5-Coder-7B-Instruct"
tokenizer = AutoTokenizer.from_pretrained(base_model_id)
base_model = AutoModelForCausalLM.from_pretrained(base_model_id, torch_dtype=torch.float16, device_map="auto")
# Load Stateless MCP Agent Adapter
model = PeftModel.from_pretrained(base_model, "beatsprom/qwen-stateless-mcp-agent-2026")
prompt = """Target Protocol: MCP 2026 Stateless JSON-RPC
Task: Execute parameterized vector similarity search on 'document_embeddings' with cosine distance. Guard against SQL injection."""
inputs = tokenizer(f"<|im_start|>user\n{prompt}<|im_end|>\n<|im_start|>assistant\n", return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
To connect this agent to live production backends, download the complete working suite on Gumroad:
.cursorrules.👉 Store Link: https://beatsprom.gumroad.com/l/stateless-mcp-agent-eval