Skip to content

Latest commit

 

History

History
179 lines (132 loc) · 7.03 KB

File metadata and controls

179 lines (132 loc) · 7.03 KB

Mefai Security Score - Methodology

Every audited project receives a Mefai Security Score from 0 to 100. This document explains exactly how that score is calculated so anyone can verify it independently.


Scoring Categories

The score is divided into 6 categories. Each category has a maximum point value and clear criteria for how points are awarded or deducted.

Category Max Points What It Measures
1. Ownership & Access Control 20 Who can modify the contract/token after deployment
2. Supply & Minting 20 Whether new tokens can be created
3. Liquidity & LP Security 20 Whether liquidity can be removed
4. Code & Program Safety 15 Contract quality, upgrade risks, known vulnerabilities
5. Fee & Transfer Mechanics 15 Hidden taxes, transfer restrictions, fee manipulation
6. Transparency & Metadata 10 Metadata integrity, verified source code, documentation

Total: 100 points


Category 1: Ownership & Access Control (20 points)

Condition Points
Ownership renounced (zero address / no authority) 20
Ownership behind multisig with timelock 16
Ownership behind multisig without timelock 12
Single owner with timelock on critical functions 8
Single owner, no timelock 4
Owner can call destructive functions (pause, blacklist, selfdestruct) 0

EVM: Checked via owner() call. Verified the return value is zero address. Solana: Checked via update authority and program upgrade authority.


Category 2: Supply & Minting (20 points)

Condition Points
No mint function / mint authority revoked 20
Mint capped at fixed max supply enforced in code 16
Mint behind multisig or governance 12
Mint behind single owner with rate limit 8
Unlimited minting by single owner 0

EVM: Source code reviewed for mint(), _mint() functions and access modifiers. Solana: Mint authority field checked via getAccountInfo. Null = revoked.


Category 3: Liquidity & LP Security (20 points)

Condition Points
LP tokens burned (permanently locked) 20
LP locked in audited locker for 12+ months 18
LP locked in audited locker for 6-12 months 16
LP locked in audited locker for 3-6 months 12
LP locked for less than 3 months 8
LP partially locked (less than 80%) 4
LP unlocked / no LP lock 0

Verification: LP lock is verified on-chain by reading the locker contract/program account directly via RPC. We decode the binary lock record to extract lock amount, lock date, and unlock date. We do not rely on third-party APIs for LP lock status.


Category 4: Code & Program Safety (15 points)

Condition Points
No vulnerabilities found, standard program/library 15
Informational findings only 13
Low severity findings only 10
Medium severity findings 6
High severity findings 3
Critical severity findings 0

EVM: Line-by-line source code review. Checked for reentrancy, overflow, access control, proxy risks, front-running, flash loan vectors. Solana: Token program type verified. Standard SPL Token = maximum safety (audited by Solana Labs). Token-2022 extensions analyzed individually. Custom programs require separate code audit.

Deductions:

  • Contract not verified on block explorer: -3
  • Uses proxy/upgrade pattern: -3
  • Compiler version below 0.8.0 (no overflow protection): -5

Category 5: Fee & Transfer Mechanics (15 points)

Condition Points
No transfer fee, no restrictions 15
Fixed fee (immutable in code, cannot be changed) 13
Fee adjustable but capped in code (e.g. max 5%) 10
Fee adjustable by owner, no cap 4
Hidden fee mechanism or reflection tax 2
Freeze/blacklist/pause authority active 0

EVM: Fee variables checked for constant or immutable modifiers. onlyOwner fee setters identified. Solana: Transfer fee extension checked. Freeze authority checked via getAccountInfo.


Category 6: Transparency & Metadata (10 points)

Condition Points
Source code verified + metadata immutable + project website active 10
Source code verified + metadata immutable 8
Source code verified but metadata mutable 6
Source code not verified but metadata immutable 4
Source code not verified and metadata mutable 2
No verifiable information 0

EVM: BscScan/Etherscan verification status. Solana: Metadata mutability flag. URI availability.


Score to Rating Conversion

Score Rating Meaning
90 - 100 LOW RISK Safe. Best practices followed. No critical concerns.
75 - 89 LOW-MEDIUM RISK Minor issues. Does not threaten funds.
60 - 74 MEDIUM RISK Attention needed. Conditional risks exist.
40 - 59 MEDIUM-HIGH RISK Significant concerns. Review before interacting.
20 - 39 HIGH RISK Direct risk to funds. Proceed with extreme caution.
0 - 19 CRITICAL RISK Immediate danger. Do not interact.

Example: MEFAI BSC Score Breakdown

Category Check Result Points
1. Ownership owner() returns zero address Renounced 20/20
2. Supply No mint() function, fixed 1B supply No minting possible 20/20
3. Liquidity LP status on PancakeSwap Verified 16/20
4. Code Safety Solidity 0.8.23, OpenZeppelin v5, 0 medium+ findings Clean 15/15
5. Fee Mechanics FEE_PERCENT is constant (1%), cannot be changed Fixed fee 13/15
6. Transparency Verified on BscScan, active website Full transparency 10/10
TOTAL 94/100

Rating: LOW RISK (94/100)


Example: MEFAI Solana Score Breakdown

Category Check Result Points
1. Ownership Mint authority = null, freeze authority = null Revoked 20/20
2. Supply Mint authority revoked, ~100M fixed supply No minting possible 20/20
3. Liquidity PinkLock until 2026-09-25 (6 months) Locked 6-12 months 16/20
4. Code Safety Standard SPL Token, no extensions, no custom program Maximum safety 15/15
5. Fee Mechanics 0% transfer fee, no freeze authority No fees, no restrictions 15/15
6. Transparency Metadata immutable, mefai.io active Full transparency 10/10
TOTAL 96/100

Rating: LOW RISK (96/100)


Principles

  1. Reproducible - Anyone can verify every score by running the same on-chain checks
  2. Objective - Criteria are binary or clearly defined, not subjective
  3. On-chain first - Every claim is backed by a blockchain query, not a third-party API
  4. Transparent - Full breakdown shown for every audited project
  5. Conservative - When in doubt, the lower score is applied

Mefai Security Research - Scoring Methodology v1.0