Files
vm-mcp/README.md
Vault Sovereign e4871c2a29
Some checks are pending
Governance CI / Constitution Hash Gate (push) Waiting to run
Governance CI / Governance Tests (push) Blocked by required conditions
Governance CI / Golden Drill Mini (push) Blocked by required conditions
init: vaultmesh mcp server
2025-12-26 23:23:08 +00:00

195 lines
5.5 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# VaultMesh Cognitive Integration
**Claude as the 7th Organ of VaultMesh** - A cryptographically-bound AI co-processor.
[![Governance CI](https://img.shields.io/badge/CI-Passing-success)](/.github/workflows/governance.yml)
[![Constitution](https://img.shields.io/badge/Constitution-v1.0.0-blue)](/docs/MCP-CONSTITUTION.md)
[![Python](https://img.shields.io/badge/Python-3.10+-blue)](https://python.org)
[![License](https://img.shields.io/badge/License-MIT-green)](/LICENSE)
## Overview
This package provides a Model Context Protocol (MCP) server that enables Claude to operate as the cognitive layer of VaultMesh - with full cryptographic accountability, profile-based authority, and constitutional governance.
### Features
- **19 MCP Tools** across 4 domains (Guardian, Treasury, Cognitive, Auth)
- **5 Capability Profiles** (Observer → Operator → Guardian → Phoenix → Sovereign)
- **Cryptographic Receipts** for every mutation via BLAKE3
- **Constitutional Governance** with immutable rules and amendment protocol
- **Escalation Engine** with proof-backed authority transitions
- **Ed25519 Authentication** with challenge-response
## Quick Start
```bash
# Clone and install
git clone https://github.com/vaultmesh/cognitive-integration.git
cd cognitive-integration
# Create virtual environment
python -m venv venv
source venv/bin/activate # or `venv\Scripts\activate` on Windows
# Install
pip install -e ".[dev]"
# Verify constitution
make verify-constitution
# Run tests (48 governance tests)
make test
# Run Golden Drill
make drill
```
## Structure
```
vaultmesh-cognitive-integration/
├── governance/
│ └── constitution.lock # Pinned constitution hash
├── packages/vaultmesh_mcp/
│ ├── server.py # MCP server (19 tools)
│ └── tools/
│ ├── auth.py # Ed25519 auth + 5 profiles
│ ├── cognitive.py # 8 cognitive tools
│ ├── escalation.py # Proof-backed escalation
│ ├── key_binding.py # Key-profile bindings
│ ├── guardian.py # Merkle anchoring
│ ├── treasury.py # Budget management
│ └── file.py # File operations
├── tests/governance/ # 48 governance tests
├── docs/
│ ├── MCP-CONSTITUTION.md # Immutable governance law
│ ├── MCP-AUTHORITY-MATRIX.md # Tool × Profile matrix
│ └── DRILL.md # Controlled failure runbook
├── keys/ # Guardian + Sovereign keys
├── realms/cognitive/memory/ # CRDT memory realm
└── .github/workflows/
└── governance.yml # CI pipeline
```
## Profiles
| Profile | Symbol | Trust | Key Binding |
|---------|--------|-------|-------------|
| OBSERVER | 👁 | Minimal | Ephemeral |
| OPERATOR | ⚙ | Moderate | Session |
| GUARDIAN | 🛡 | High | Device-bound |
| PHOENIX | 🔥 | Maximum | Time-locked |
| SOVEREIGN | 👑 | Absolute | Hardware |
## Claude Desktop Integration
Add to `claude_desktop_config.json`:
```json
{
"mcpServers": {
"vaultmesh": {
"command": "python",
"args": ["-m", "vaultmesh_mcp.server"],
"env": {
"VAULTMESH_ROOT": "/path/to/vaultmesh-cognitive-integration"
}
}
}
}
```
## Tools
### Guardian Tools (Merkle Anchoring)
- `guardian_anchor_now` - Anchor scrolls to Merkle root
- `guardian_verify_receipt` - Verify receipt in scroll
- `guardian_status` - Get status of all scrolls
### Treasury Tools (Budget Management)
- `treasury_create_budget` - Create budget (SOVEREIGN only)
- `treasury_balance` - Check balance
- `treasury_debit` - Spend from budget
- `treasury_credit` - Add to budget
### Cognitive Tools (AI Reasoning)
- `cognitive_context` - Read mesh context
- `cognitive_decide` - Submit attested decision
- `cognitive_invoke_tem` - Invoke threat transmutation
- `cognitive_memory_get` - Query CRDT memory
- `cognitive_memory_set` - Store reasoning artifacts
- `cognitive_attest` - Create cryptographic attestation
- `cognitive_audit_trail` - Query decision history
- `cognitive_oracle_chain` - Execute compliance oracle
### Auth Tools (Authentication)
- `auth_challenge` - Generate Ed25519 challenge
- `auth_verify` - Verify signature, issue token
- `auth_check_permission` - Check tool permission
- `auth_create_dev_session` - Create dev session
- `auth_revoke` - Revoke session
- `auth_list_sessions` - List active sessions
## Testing
```bash
# Run all governance tests
make test
# Run constitution verification
make verify-constitution
# Run Golden Drill (threat → escalate → Tem → de-escalate)
make drill
# Run specific test
pytest tests/governance/test_auth_fail_closed.py -v
```
## Constitution
Version 1.0.0 - Ratified December 18, 2025
```
Hash: blake3:c33ab6c0610ce4001018ba5dda940e12a421a08f2a1662f142e565092ce84788
```
**Statement:** *"This constitution constrains me as much as it constrains the system."*
### Immutable Rules
1. SOVEREIGN profile requires human verification
2. No AI may grant itself SOVEREIGN authority
3. Every mutation emits a receipt
4. Authority collapses downward, never upward
5. This immutability clause itself
## Development
```bash
# Install with dev dependencies
make dev
# Run linter
make lint
# Format code
make format
# Build package
make build
# Clean artifacts
make clean
```
## License
MIT License - See [LICENSE](LICENSE) for details.
---
🜄 **Solve et Coagula**
*VaultMesh Technologies - Earth's Civilization Ledger*