|
--- |
|
language: |
|
- en |
|
license: apache-2.0 |
|
tags: |
|
- Aptos |
|
- Move |
|
- Llama3 |
|
- Blockchain |
|
- SmartContract |
|
--- |
|
|
|
# Uploaded model |
|
- **Name**: L3_8B_Aptos_Smart_Contract_Analyser |
|
- **Organization**: Armur |
|
- **Project**: Aptos Smart Contract Audit |
|
- **License**: apache-2.0 |
|
- **language**: en |
|
|
|
# Usage |
|
|
|
```python |
|
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer |
|
|
|
model_path = "PATH_TO_THIS_REPO" |
|
|
|
tokenizer = AutoTokenizer.from_pretrained(model_path) |
|
model = AutoModelForCausalLM.from_pretrained( |
|
model_path, |
|
device_map="auto", |
|
torch_dtype='auto' |
|
).eval() |
|
|
|
# Prompt content: "hi" |
|
messages = [ |
|
{"role": "user", "content": Move_Smart_Contract} |
|
] |
|
|
|
input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt') |
|
output_ids = model.generate(input_ids.to('cuda')) |
|
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True) |
|
|
|
# Model response: "-- Vulnerabilites --" |
|
print(response) |
|
``` |
|
|
|
# L3_8B_Aptos_Smart_Contract_Analyser |
|
|
|
L3_8B_Aptos_Smart_Contract_Analyser is a cutting-edge AI model designed for comprehensive analysis and security auditing of smart contracts on the Aptos blockchain. We utilized the `ArmurAI/L3_8B` base model, enhanced it with extensive knowledge of Aptos-specific vulnerabilities and best practices, and fine-tuned it with a vast dataset of both secure and vulnerable smart contracts. |
|
|
|
# Key Features of L3_8B_Aptos_Smart_Contract_Analyser: |
|
|
|
## Advanced Vulnerability Detection |
|
|
|
Identifies critical security flaws in Aptos smart contracts: |
|
|
|
- Reentrancy vulnerabilities specific to Aptos' resource-oriented architecture |
|
- Unauthorized access to sensitive functions or resources |
|
- Integer overflow/underflow issues in financial calculations |
|
- Improper error handling that could lead to unexpected contract behavior |
|
- Resource exhaustion attacks unique to Aptos' gas model |
|
|
|
## Move-Specific Security Analysis |
|
|
|
Leverages deep understanding of the Move language to detect: |
|
|
|
- Incorrect usage of Move's ownership and borrowing rules |
|
- Potential violations of linear logic principles in resource management |
|
- Unsafe type casting or conversions that could compromise contract integrity |
|
|
|
## Aptos Module Interaction Auditing |
|
|
|
Identifies risks in inter-module interactions: |
|
|
|
- Unauthorized calls to privileged functions in other modules |
|
- Potential front-running vulnerabilities in multi-step operations |
|
- Inconsistent state management across module boundaries |
|
|
|
## Tokenomics and Access Control Verification |
|
|
|
Scrutinizes: |
|
|
|
- Flaws in token distribution mechanisms that could lead to unfair advantages |
|
- Weaknesses in role-based access control implementations |
|
- Potential centralization risks in governance structures |
|
|
|
## Smart Contract Upgrade Vulnerabilities |
|
|
|
Detects: |
|
|
|
- Insecure upgrade patterns that could lead to loss of funds or contract takeover |
|
- Inconsistencies between different versions of upgraded contracts |
|
- Potential backdoors or hidden admin functions in upgradeable contracts |
|
By incorporating these features, L3_8B_Aptos_Smart_Contract_Analyser not only enhances the security of your Aptos smart contracts but also improves their efficiency and quality. Whether you're a seasoned Aptos developer or new to the ecosystem, our model provides invaluable support throughout your development lifecycle. |
|
|
|
For more details and to access the tool, visit the [ArmurAI GitHub repository](https://github.com/Armur-Ai/Aptos-Smart-Contract-Auditor). |
|
|
