mcpField
V
Vigil - AI Security Audit and Cryptographic Proof

Vigil - AI Security Audit and Cryptographic Proof

Local
github.com/vigil-xyDeveloper Tools6Updated 32w ago
Repository

About Vigil - AI Security Audit and Cryptographic Proof

Vigil is a dashboard for teams managing AI agents, with controls for assignments, budgets, live activity, alerts, and compliance reporting. It connects OpenAI, Anthropic, and custom agents.

Key features

  1. 01Connect OpenAI, Anthropic, and custom agents
  2. 02Assign tasks, goals, and budgets
  3. 03Monitor agent activity and costs
  4. 04Send alerts for policy issues
  5. 05Export compliance audit trails

Use cases

  1. 01Set spending limits for agent tasks
  2. 02Track agent activity across teams
  3. 03Alert on unauthorized tools or PII
  4. 04Prepare SOC 2 or ISO 42001 reports
#security#vulnerabilities#ai-system

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "vigil": {
      "command": "npx",
      "args": [
        "-y",
        "vygil"
      ]
    }
  }
}
Run commandnpx -y vygil

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmvygil

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

Control your AI agents. Track costs. Stay compliant. open-souce

What It Does

One dashboard to manage your AI team.

FeatureHow It Helps
Connect your agentsPlug in OpenAI, Anthropic, or custom agents
Set goals & budgetsAssign work, cap costs, define success
Watch in real-timeSee what every agent is doing right now
Get alertsSlack, email, or PagerDuty when something's off
Prove complianceExport audit trails for SOC 2, ISO 42001

Quick Start (5 Minutes)

1. Start the database

docker compose up -d postgres

2. Install & configure

npm install
cp apps/api/.env.example apps/api/.env
cp apps/web/.env.example apps/web/.env

Or run one command:

npm run setup:local

3. Run it

# Terminal 1: API
npm run dev:api

# Terminal 2: Dashboard
npm run dev:web

Or run both in one terminal:

npm run dev:all

4. Open http://localhost:5173

Verify Everything Is Connected
npm run check:stack

If API + web are running, seed demo data:

npm run seed:demo

Troubleshooting (macOS)

1) cp .env.example ... fails

Use the correct paths from repo root:

cp apps/api/.env.example apps/api/.env
cp apps/web/.env.example apps/web/.env
2) Cannot connect to the Docker daemon
  • Start Docker Desktop
  • Wait until Docker says it is running
  • Retry:
docker compose up -d postgres
3) API error ECONNREFUSED ... :5432

PostgreSQL is not reachable. Confirm:

docker compose ps

Then restart API:

npm run dev:api

Example: Customer Support Team

Create a team

POST /api/teams
{
  "name": "Support Squad",
  "agents": [
    { "provider": "openai", "model": "gpt-4", "role": "triage" },
    { "provider": "anthropic", "model": "claude-3", "role": "escalation" }
  ],
  "goals": [
    { "metric": "response_time", "target": "< 2 min" }
  ]
}

Assign work

POST /api/tasks
{
  "team_id": "<your-team-id>",
  "description": "Handle refund requests",
  "budget": { "usd": 5.00, "tokens": 10000 },
  "policies": ["no_pii_exposure", "enforce_budget_limits"]
}

Get alerts when agents go off-script

  • Budget exceeded -> Slack notification
  • Unauthorized tool -> Email alert + auto-stop
  • PII detected -> PagerDuty escalation

Export compliance reports

GET /api/compliance/report?type=SOC_2_Type_II

Dashboard

ScreenWhat You See
OverviewAll teams, active agents, cost burn today
AgentsAdd/edit teams, connect providers
TasksQueue, assign work, monitor progress
CostsBudgets, spend trends, forecasts
GovernancePolicy violations, audit logs
AlertsReal-time feed, escalation history

Built With

  • Backend: Node.js + Express + PostgreSQL
  • Frontend: React + Vite
  • Integrations: OpenAI, Anthropic, Slack, PagerDuty, email

Ready to control your AI team?

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.