mcpField
D
DCL Evaluator

DCL Evaluator

Local
github.com/Fronesis-LabsAI & Agents0Updated 2w ago
Repository

About DCL Evaluator

DCL Evaluator is an earlier product for creating tamper-evident cryptographic audit trails of AI agent decisions. Its README marks it as superseded and directs agents to the hosted DCL Trust Oracle; the earlier MCP endpoint and tools are historical.

Key features

  1. 01SHA-256 hash-chain records
  2. 02Merkle-tree audit trails
  3. 03Deterministic policy evaluation
  4. 04Statistical behavioural-drift detection
  5. 05Compliance report generation
  6. 06Offline operation

Use cases

  1. 01Record agent actions for later integrity checks
  2. 02Review decision chains for tampering
  3. 03Detect changes in agent behaviour
  4. 04Generate audit reports for compliance review

MCP support

The product's former MCP server exposed tools to commit actions, verify chain integrity, retrieve audit trails, and generate reports. The README says this endpoint is superseded and points to the hosted DCL Trust Oracle at https://mcp.fronesislabs.com/mcp.

#audit-cryptography-ai-agents-llm-compliance

Configuration

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

mcp.json · Universal config
{
  "mcpServers": {
    "dcl-evaluator": {
      "command": "npx",
      "args": [
        "-y",
        "@smithery/cli"
      ],
      "env": {
        "DCL_API_KEY": "<DCL_API_KEY>"
      }
    }
  }
}
Run commandnpx -y @smithery/cli
Environment variables to set
DCL_API_KEYSecret

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
npm@smithery/cli

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

Superseded. DCL Evaluator is the earlier DCL product, including its MCP endpoint (/sse, API key) and its offline desktop audit trail. Agents should use the hosted DCL Trust Oracle at https://mcp.fronesislabs.com/mcp. References below to /sse, API keys, dcl_commit, dcl_verify, dcl_get_chain, dcl_report, and fixed prices are historical.

DCL Evaluator

Version License Patent Platform MCP Offline LLMs dcl-evaluator MCP server

Cryptographic audit trail for AI agents. Powered by Leibniz Layer™

Every decision your AI agent makes — sealed, verified, auditable.

DCL Evaluator is the first implementation of Leibniz Layer™ — a cryptographic verification protocol for AI agent decisions. It brings deterministic, tamper-evident audit trails to any LLM-powered system.

Why DCL Evaluator

AI agents make decisions. Those decisions carry risk — legal, financial, reputational. Yet most AI systems are black boxes: no record of what was decided, why, or whether the decision was tampered with afterward.

DCL Evaluator solves this with a simple principle borrowed from cryptography: commit every decision into a hash chain. Modify any past record and the entire chain invalidates. Mathematical proof of integrity — no trust required.

Core capabilities

  • SHA-256 hash chain — every agent action cryptographically sealed
  • Merkle tree audit trail — tamper-evident by design
  • Deterministic policy engine — identical input = identical decision, 100% reproducible
  • Drift detection — statistical Z-test catches behavioural drift before it becomes a compliance failure
  • Multi-LLM support — Claude, GPT-4, Grok, Gemini, DeepSeek, Ollama (local, air-gapped)
  • MCP server — connect any agent in minutes via Model Context Protocol
  • Compliance reports — tamper-evident PDF export, ready for regulators
  • Offline-capable — runs fully air-gapped, zero data leaves your infrastructure

💳 Pay-per-Audit via x402

DCL Evaluator supports the x402 payment protocol.
AI agents can request audits autonomously — no API keys, just HTTP 402 + USDC on Base.

Endpoint: https://x402.fronesislabs.com
Network: Base Mainnet (eip155:8453)

ToolPriceMargin
Quick Scan$1 USDC99%
Standard Audit (4 models)$5 USDC99.4%
Deep Audit + PoC$15 USDC99.3%

Any agent that speaks HTTP can pay for and receive a cryptographic audit — no human required.

Connect via MCP

{
  "mcpServers": {
    "dcl-evaluator": {
      "url": "https://mcp.fronesislabs.com/sse",
      "headers": {
        "x-api-key": "your-key"
      }
    }
  }
}

Four tools available:

ToolDescription
dcl_commitSeal an agent action into the hash chain
dcl_verifyVerify chain integrity
dcl_get_chainRetrieve full audit trail
dcl_reportGenerate compliance report

Health check: https://mcp.fronesislabs.com/health

Download

→ Latest release — v1.2.0

Desktop app (Windows). Wails + Go + React. Works offline.

Compliance coverage

Built-in policy templates:

DEFAULT · EU AI ACT · GDPR · FINANCE · MEDICAL · RED TEAM

Leibniz Layer™

DCL Evaluator is the first product implementing Leibniz Layer™ — the cryptographic verification layer for the AI agent economy.

Named after Gottfried Wilhelm Leibniz — mathematician, logician, pioneer of deterministic reasoning.

"Every decision, every action deterministically sealed, tamper-evident, auditable."

leibniz.fronesislabs.com

Links

© 2026 Fronesis Labs · Leibniz Layer™, DCL Evaluator™ — Patent Pending · Source-Available License

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.