alt-research/solidityguard/openclaw-skill

Claude Code and agent skill by alt-research/solidityguard

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Advanced Solidity/EVM smart contract security auditor with 104 vulnerability patterns, multi-tool integration, and professional report generation.

Install

npx vskill@latest install alt-research/solidityguard/openclaw-skill

SKILL.md

SolidityGuard — Smart Contract Security Auditor

Advanced Solidity/EVM smart contract security audit skill with 104 vulnerability patterns covering reentrancy, access control, arithmetic, DeFi, proxy, oracle, transient storage (EIP-1153), account abstraction (ERC-4337), and EIP-7702/Pectra.

When to Use

Use this skill when:

  • The user asks to audit, scan, or review a Solidity smart contract
  • The user shares .sol or .vy files and asks about security
  • The user mentions vulnerability, exploit, reentrancy, access control, or security audit
  • The user asks about smart contract best practices or common attack vectors
  • The user wants to generate a security report for their contracts
  • The user wants fuzz tests (Foundry/Echidna) generated for their contracts
  • The user asks about specific vulnerability patterns (e.g., "check for flash loan attacks")

Tools

  • solidityguard_scan: Scans Solidity files for 104 vulnerability patterns using 50+ detectors
  • solidityguard_report: Generates a professional audit report (Markdown/PDF) from scan findings
  • solidityguard_fuzz: Generates Foundry invariant tests and Echidna property-based fuzz tests

Instructions

Quick Scan

To scan a single contract or directory for vulnerabilities:

bash
python3 scripts/solidity_guard.py scan <path-to-contracts>

This runs the pattern scanner with 50+ detectors covering all 104 vulnerability patterns. Output is JSON with findings including severity, confidence, file:line location, and remediation.

Full Audit Workflow

For a comprehensive audit, follow this 7-phase workflow:

Phase 1 — Entry Point Analysis Enumerate all external/public functions to map the attack surface:

bash
# Find all entry points
grep -rn "function.*external\|function.*public" contracts/

Phase 2 — Automated Scan Run available security tools, then the pattern scanner:

bash
# Slither (if installed)
slither . --json slither-results.json 2>/dev/null || true
# Aderyn (if installed)
aderyn -s contracts/ -o aderyn-report.md 2>/dev/null || true
# Pattern scanner (always available)
python3 scripts/solidity_guard.py scan contracts/ --output findings.json

Phase 3 — Finding Verification Cross-reference tool outputs. Findings confirmed by 2+ tools get a confidence boost (+10%). Findings confirmed by 3+ tools are capped at 95% confidence.

Phase 4 — Deduplication Remove duplicate findings across tools. Keep the highest-confidence instance.

Phase 5 — Confidence Filtering Filter out findings below 0.70 confidence threshold.

Phase 6 — Report Generation Generate a professional audit report:

bash
python3 scripts/report_generator.py --input findings.json --output audit-report.md

Phase 7 — Remediation For each finding, provide:

  • Exact file:line location with vulnerable code snippet
  • Step-by-step attack scenario
  • Fixed Solidity code
Vulnerability Categories

The scanner detects 104 patterns organized into these categories:

CategoryPatternsExamples
ReentrancyETH-001 to ETH-005Single, cross-function, cross-contract, read-only, cross-chain
Access ControlETH-006 to ETH-012Missing auth, tx.origin, selfdestruct, proxy, centralization
ArithmeticETH-013 to ETH-017Overflow, division-before-multiply, unchecked, rounding, precision
External CallsETH-018 to ETH-023Unchecked return, delegatecall, low-level call, DoS, gas griefing
Oracle/PriceETH-024 to ETH-028Manipulation, flash loan, MEV/sandwich, slippage, staleness
Storage/StateETH-029 to ETH-033Uninitialized pointer, collision, shadowing, unexpected ether
Logic ErrorsETH-034 to ETH-040Strict equality, TOD, timestamp, randomness, signature, front-run
Token IssuesETH-041 to ETH-048Non-standard ERC20, fee-on-transfer, rebasing, ERC777, zero-addr
Proxy/UpgradeETH-049 to ETH-054Uninitialized impl, storage mismatch, selector clash, upgrade auth
DeFiETH-055 to ETH-065Governance, liquidation, vault inflation, AMM, flash mint, reward
Gas/DoSETH-066 to ETH-070Unbounded loop, block gas limit, revert-in-loop, griefing
MiscellaneousETH-071 to ETH-080Floating pragma, compiler, encodePacked, RTLO, inheritance
Transient StorageETH-081 to ETH-085Slot collision, not cleared, TSTORE reentry bypass, delegatecall
EIP-7702/PectraETH-086 to ETH-089Broken EOA check, delegation, cross-chain replay, code assumption
ERC-4337 AAETH-090 to ETH-093UserOp collision, paymaster, bundler, validation-execution
Modern DeFiETH-094 to ETH-097Uniswap V4 hooks, hook data, cached state, compiler bugs
Input ValidationETH-098 to ETH-099Boundary checks, unsafe ABI decoding
Off-ChainETH-100 to ETH-101Delegation phishing, infrastructure compromise
Restaking/L2ETH-102 to ETH-104Cascading slashing, sequencer dependency, message replay
Finding Format

Every finding MUST include:

## [SEVERITY] ETH-XXX: Vulnerability Name
**Location**: `contracts/File.sol:123` (functionName)
**Confidence**: 0.XX
**Category**: Category Name
### Description
What the vulnerability is and why it matters.
### Evidence

// Vulnerable code from contracts/File.sol:123 <exact code snippet>

### Attack Scenario
1. Attacker calls functionA()
2. During callback, attacker re-enters functionB()
3. State is inconsistent, allowing double withdrawal
### Recommendation

// Fixed code <secure implementation>

Anti-Hallucination Rules

CRITICAL — All findings MUST include:

  • Exact file:line code location (not approximated)
  • Vulnerable code snippet (verbatim from source, not generated)
  • Specific attack scenario with numbered steps
  • Remediation with working Solidity code
  • Confidence score >= 0.70

REJECT any finding that:

  • Says "this looks vulnerable" without exact pattern match
  • Says "probably missing check" without verifying all code paths
  • Has confidence below 0.70
  • Cannot cite the exact line of vulnerable code
Tool Integration

The skill works best with these optional tools installed:

ToolInstallPurpose
Slitherpip install slither-analyzerPrimary static analyzer
Aderyn`curl -L https://raw.githubusercontent.com/Cyfrin/aderyn/dev/cyfrinup/installbash && cyfrinup`Fast Rust-based analyzer
Mythrilpip install mythrilSymbolic execution
Foundry`curl -L https://foundry.paradigm.xyz \bash && foundryup`Testing + fuzzing
Benchmarks

Validated against real-world CTF challenges:

BenchmarkContractsDetection
DeFiVulnLabs56 contracts, 59 patterns100%
Paradigm CTF 202110 static challenges100%
Paradigm CTF 20227 static challenges100%
Paradigm CTF 20237 static challenges100%
OWASP Smart Contract Top 10 (2025) Coverage
RankCategoryPatterns
#1Access ControlETH-006 to ETH-012, ETH-049 to ETH-054
#2Oracle ManipulationETH-024 to ETH-028
#3Logic ErrorsETH-034 to ETH-040
#4Input ValidationETH-098, ETH-099
#5ReentrancyETH-001 to ETH-005
#6Unchecked ReturnsETH-018 to ETH-023
#7MEV / Front-runningETH-026, ETH-040, ETH-060
#8ArithmeticETH-013 to ETH-017
#9Unsafe DelegatecallETH-019, ETH-030
#10Denial of ServiceETH-066 to ETH-070

Examples

  • "Audit my contracts for reentrancy" → Runs reentrancy-focused scan (ETH-001 to ETH-005)
  • "Scan this DeFi protocol" → Full 104-pattern scan with all available tools
  • "Check for flash loan vulnerabilities" → Targeted scan for ETH-024, ETH-025, ETH-057
  • "Generate a security report" → Professional Markdown audit report
  • "Write fuzz tests for this vault" → Foundry invariant tests + Echidna properties
  • "Is this proxy upgradeable safely?" → Storage layout + proxy pattern analysis

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
81
Repository forks
19
7-day trend score
22

Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 18 Mar 2026, 7 months ago.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/alt-research/solidityguard/openclaw-skill/badge)](https://verified-skill.com/skills/alt-research/solidityguard/openclaw-skill)

Something wrong with this skill or its page? Tell us.