Search

Prompt injection detection for LLM applications and MCP servers

Agent-native incident investigation graph: analysts and AI agents (MCP/REST) link evidence from EDR, Entra ID, cloud and GitHub logs into one shared, real-time attack graph and timeline. Self-hosted, no database.

Temporary package placeholder for staged publishing

MCP server exposing the OAK (OnChain Attack Knowledge) framework as queryable tools for AI coding agents.

Temporary package placeholder for staged publishing

An MCP server that validates tool calls against JSON Schema, performs deterministic repair, redacts secrets, and maintains a hash-chained audit ledger.

Runtime + semantic trust registry for MCP servers - catches tool poisoning and scope creep, $0 infra (local LLM via Ollama)

Dual LLM with Symbolic Memory: a reference-monitor middleware that neutralises indirect prompt injection and MCP tool poisoning by design (Python, mypy --strict).

MCP server fuzzer client and utilities

Python SDK and MCP server for Polaxis - AI agent governance platform. Evaluate every tool call against policies in real time: block dangerous actions, enforce budgets, route to human approvals. Under 5ms.

Local-first verification layer for AI agents: evidence, provenance, policy gates, human approval, and auditable Trust Receipts before a claim, memory write, or action is trusted.

Zero-Trust Data Sanitization (ZTDS RFC v1.0) reference MCP server for Claude Desktop, Cursor, and Windsurf. In-memory PII de-identification and deterministic surrogate tokenization.

Deterministic MCP proxy that blocks money-moving and data-egress tool calls in any agent session that has received untrusted content. No model in the decision path.

MCP server and CLI that scan untrusted content for prompt-injection patterns before it reaches your agent - offline, explainable, cross-agent.

Aegis Jail MCP: least-privilege sandbox for AI coding agents and MCP servers. Hides ~/.ssh, secrets and the network, proves the walls hold, finds the minimal permissions. Linux, macOS, Windows. Claude Code plugin + MCP server.

MCP Proxy — runtime interception for AI agent tool calls using ATR rules

Organize all your Claude Code memories, skills, MCP servers, and hooks — view by scope hierarchy, move between scopes via drag-and-drop

Test regular expressions in a timeout-guarded worker, explain them, find catastrophic backtracking.

Enforcing proxy for MCP tool calls: every tools/call is checked against an FDIA policy before it reaches the server, with a hash-chained audit log.

A vendor-neutral collection of Sigma, YARA, and Suricata rules for detecting compromise of LLM applications, MCP servers, ML supply chains, AI infrastructure, AI-powered insider threats, and RAG/vector database attacks.