tardigrade-shield
Model-agnostic cybersecurity MCP server for CMMC 2.0 compliance workflows with air-gapped Ollama mode.
README
tardigrade-shield
Model-agnostic cybersecurity MCP (Model Context Protocol) server built for CMMC 2.0 compliance workflows, with a local / air-gapped Ollama mode for environments that can't touch the network.
Status: in progress — nine-file Claude Code build package already assembled (see that pack for implementation task tracking; this README is the public-facing project description).
What this is
A security-tooling layer exposed as MCP tools, so any MCP-compatible model (or the model-agnostic local mode) can be given controlled, auditable access to compliance-relevant checks — rather than each check being hand-run or hand-coded per environment.
Design principles
- Model-agnostic: the MCP interface is the contract, not any particular model's capabilities. Swapping the underlying model (cloud-hosted or local) shouldn't require rewriting the tool definitions.
- Air-gapped mode: a local Ollama-backed mode for environments where no outbound network access is permitted at all — a real constraint in CMMC 2.0 contexts, not a nice-to-have.
- Auditable by construction: every tool call should produce a record suitable for a compliance audit trail, not just a chat response.
Scope: CMMC 2.0
CMMC 2.0 organizes controls into families (access control, incident response, system integrity, etc.). This project does not attempt full coverage in v1 — see Open Questions for which families are prioritized.
Repo structure
tardigrade-shield/
├── src/
│ ├── mcp_server/ # MCP tool definitions and server entrypoint
│ ├── controls/ # one module per CMMC 2.0 control family covered
│ ├── ollama_mode/ # local/air-gapped model backend
│ └── audit/ # audit trail generation for compliance review
└── tests/
Implementation details and task-by-task build tracking live in the nine-file Claude Code build package assembled separately — this repo should end up matching that pack's structure once built out.
Status / roadmap
- [ ] Execute the nine-file build package
- [ ] Validate against a representative CMMC 2.0 control set
- [ ] Air-gapped mode smoke test with a local Ollama model
Open questions
- Which specific CMMC 2.0 control families are covered first vs. deferred to a later version — not yet decided
- Whether the model-agnostic design needs a compatibility shim per model family beyond Ollama
Requirements
Node ≥ 20 or Python ≥ 3.10 (implementation language not yet finalized — see the build package for the current decision). Ollama for local mode.
License
MIT — see LICENSE.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.