AMEVA MCP Advanced Toolkit
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AMEVA MCP Advanced Toolkit scan this source file for hardcoded secrets"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
AMEVA MCP Advanced Toolkit
This repository provides a set of high-performance, WebAssembly (WASM) based Model Context Protocol (MCP) tools designed to enable secure, sandboxed, and zero-host-pollution operations directly within local runtime environments.
All operations are executed within the virtual memory of WASM runtimes compiled from Rust and C, ensuring zero-footprint local processing without host pollution or external network dependency.
Detailed tool-specific specifications, API schemas, and interface structures are available in the WASM Engines Specification (Local File Link) and WASM Engines Specification (GitHub Relative Link).
Tool Suite Overview
# | Tool Name | Engine/Language | Functional Description |
1 |
| C (QuickJS) | Zero-dependency, isolated JavaScript execution environment. |
2 |
| C (Capstone) | Local binary disassembler for x86/ARM machine code analysis. |
3 |
| Rust | Raw byte structure parsing and reconstruction for damaged ZIP/PNG/JPEG/SQLite files. |
4 |
| Rust | Least Significant Bit (LSB) steganographic encoder. |
5 |
| Rust | Steganographic decoder to extract hidden payload from images. |
6 |
| Rust | High-speed, local image compression and format conversion (WebP/AVIF). |
7 |
| Rust (Polars) | In-memory massive CSV/DataFrame query engine with SQL support. |
8 |
| Rust | Secure personal identifiable information (PII) masking tool for PDF documents. |
9 |
| Rust | Static analysis scanner for hardcoded API keys, tokens, and credentials. |
10 |
| Rust (ONNX) | In-browser semantic vector similarity search via local ONNX embeddings. |
Related MCP server: Code Context Provider MCP
Build System
Prerequisites
Rust and target
wasm32-unknown-unknownNode.js 18 or higher
Build Environment Setup
rustup target add wasm32-unknown-unknown
npm installComplete Compilation
npm run build:allMCP Integration Configuration
To integrate this toolkit into Cursor, Claude Desktop, or other MCP clients, add the following configuration to your mcpServers profile:
{
"mcpServers": {
"ameva-advanced": {
"command": "node",
"args": ["C:/ameva/AMEVA-MCP-Advanced-Toolkit/src/proxy.js"]
}
}
}Directory Structure
mcp_manifest.json : System-wide MCP tool schema definitions.
package.json : Build automation pipeline and dependencies.
src/proxy.js : JSON-RPC bridge handler bridging MCP clients to WASM modules.
wasm/ : Sub-directory housing the source code and build workflows of the WASM engines.
Security Principles
Zero Host Pollution: No temporary or persistent files are written to the host filesystem during operations.
Offline First: All computations are performed strictly inside the runtime's memory layout; network traffic is disabled.
Sandboxed Execution: Virtual execution domains isolate operations, mitigating host-level security vulnerabilities.
AMEVA Universe — uno-km
Available Tools
10 toolsdisassemble_binaryA
의심스러운 .exe나 기계어 바이너리 파일을 호스트 PC에서 직접 실행하지 않고 WASM 격리 환경에서 x86_64/ARM64 어셈블리로 디스어셈블하여 분석합니다.
Inputs: binary_base64 (required), arch (required), mode.
execute_sandboxA
Docker 설치 없이 브라우저 WASM 메모리 내 격리된 가상 환경에서 JavaScript 코드를 안전하게 실행하고 즉시 소멸시킵니다. 파일시스템/네트워크 접근이 완전히 차단됩니다.
Inputs: code (required), timeout_ms.
optimize_imagesB
PNG/JPEG 이미지를 외부 서버로 전송하지 않고 100% 로컬 WASM 환경에서 무손실/유손실 압축하고 WebP/AVIF 등 원하는 포맷으로 일괄 변환합니다.
Inputs: images (required), target_format, quality, lossless.
process_dataframeA
수백만 줄의 CSV/TSV 데이터를 Rust 기반 Polars 엔진으로 브라우저 내에서 초고속 처리합니다. 필터링, 정렬, 집계, JOIN 등 SQL 쿼리를 지원합니다.
Inputs: csv_data (required), query (required), output_format.
recover_fileA
손상된 ZIP, PNG, JPEG, SQLite 파일의 원시 바이트 구조를 Rust 복구 엔진으로 역추적하여 복구합니다. 파일 시그니처를 재건하고 가능한 데이터를 카빙합니다.
Inputs: file_base64 (required), file_type, strategy.
redact_pdfA
대외비 PDF 문서에서 주민번호, 계좌번호, 이메일, 전화번호 등 개인정보를 오프라인 WASM 환경에서 완전히 마스킹한 새 PDF를 생성합니다. 외부 서버로 1바이트도 전송하지 않습니다.
Inputs: pdf_base64 (required), patterns, custom_regex, mode.
scan_secretsA
소스코드 파일이나 디렉토리를 커밋 전에 스캔하여 하드코딩된 AWS/GCP/Slack API 키, DB 비밀번호, JWT 토큰 등을 탐지하고 파일명 및 라인 번호를 포함한 경고 보고서를 생성합니다.
Inputs: files (required), entropy_threshold, output_format.
semantic_searchA
소스코드나 문서에서 자연어 질문으로 의미 기반(Semantic) 검색을 수행합니다. ONNX 임베딩 모델을 WASM으로 구동하여 외부 API 없이 로컬에서 벡터 유사도 검색합니다.
Inputs: query (required), documents (required), top_k.
steg_decodeB
스테가노그래피로 인코딩된 이미지에서 숨겨진 기밀 데이터를 추출합니다.
Inputs: steg_image_base64 (required), password.
steg_encodeA
임의의 기밀 텍스트나 바이너리 데이터를 PNG/BMP 이미지의 픽셀 LSB에 육안으로 식별 불가능하게 숨깁니다 (스테가노그래피 인코딩).
Inputs: cover_image_base64 (required), secret_data (required), password.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Each tool has a clearly distinct purpose: binary disassembly, sandboxed JS execution, image optimization, dataframe processing, file recovery, PDF redaction, secret scanning, semantic search, and steganography encode/decode. No two tools overlap in functionality, making selection unambiguous.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., disassemble_binary, execute_sandbox, steg_decode). The naming convention is uniform and predictable.
With 10 tools, the set is well-scoped for an 'Advanced Toolkit' covering security, data processing, and file analysis. Each tool serves a distinct and useful purpose, and the count is neither too sparse nor overwhelming.
The tool surface covers a broad range of security and data processing tasks (binary analysis, sandboxed execution, image optimization, dataframe processing, file recovery, PDF redaction, secret scanning, semantic search, steganography). Minor gaps exist, such as the absence of general encryption/decryption or network analysis tools, but the set is comprehensive for its intended scope.
Related MCP Connectors
84+ free local-first tools: image, PDF, docs, dev utils. Wasm, zero upload, x402 API.
Execution platform for AI agents on Firecracker microVMs in the EU. Interactive Linux sandboxes with root, templates for Python, Node.js, Go, Rust and a headless browser, public HTTPS URLs. Background jobs that outlive the connection, batches of up to 1,000 jobs with dependencies, hard cost caps reserved up front, execution receipts with SHA-256, log digests that fit the context. IPv4, dual-stack or IPv6-only with its own public address, allowlists or no network. WASI runtime for small tasks. Idle sandboxes sleep and keep their files. Also REST API, Python SDK and LangChain.
144 deterministic file tools: PDF, image, media, convert, analyze. Connect in one click (OAuth).
26 deterministic tools: PDF, image, QR, CSV/Excel, HTML to PDF, DNS/RDAP, FX rates, time zones.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceProvides secure execution of arbitrary JavaScript code within a sandboxed QuickJS WASM environment, allowing language models or other MCP clients to safely run JavaScript code snippets without compromising the host system.5-
- AlicenseAqualityDmaintenanceProvides code context and analysis for AI assistants by extracting directory structures and code symbols using WebAssembly Tree-sitter parsers with zero native dependencies.120MIT
- AlicenseNot gradedqualityDmaintenanceProvides secure filesystem operations, HTTP fetching with SSRF protection, JSON validation, artifact logging, and optional Redis key-value storage through both stdio and HTTP transports. Features production-ready security controls including sandbox enforcement, allowlist validation, and comprehensive input validation.Apache 2.0
- AlicenseNot gradedqualityNot gradedmaintenanceExposes common CTF and cybersecurity tools (crypto, forensics, malware analysis, steganography, reverse engineering, pwn, OSINT) so LLMs can help solve capture-the-flag challenges in a controlled lab environment.-