A highly efficient, collaborative agentic coding framework using Model Context Protocol (MCP) to divide work between planning, code graph analysis, and execution.
专为 OpenClaw 等 AI 智能体解决失忆、不自主进化的记忆成长系统，让您的智能体越用越聪明。
Official Model Context Protocol (MCP) server for local-first PDF/A generation. Empowers AI agents (Claude, Cursor, Zed) to securely create, sign (PAdES), verify, and transform documents with barcodes and AcroForms.
agent-sandbox ai-agents windows sandbox mxc microsoft-execution-container processcontainer typescript nodejs cursor claude-code agent-sandboxing wsl2 npm sdk security blast-radius
This Python application scans for the CVE-2023-38831 vulnerability in WinRAR.
Governed runtime for autonomous AI agents with capability permissions, policy enforcement, and sandboxed execution.
Quantum Field Theory security framework using Feynman diagrams for dynamic ISO management. Implements physics-based isolation where attacks cause decoherence, not direct compromise. Three-state architecture: Closed (Ụzọ Mechiri), Open (Ụzọ Mepere), Protected (Ụzọ Nchekwa). When quantum systems fail, build your own.
A security engineer's playbook for defending agentic AI coding assistants (Claude Code, Cowork, Codex, Cursor) across laptops, remote workspaces, and GPU/sandbox clouds — structured as an LLM Wiki.
Add a description, image, and links to the sandbox-security topic page so that developers can more easily learn about it.
To associate your repository with the sandbox-security topic, visit your repo's landing page and select "manage topics."