Package Exports

@ezark-publish/agentdesk-mcp
@ezark-publish/agentdesk-mcp/dist/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (@ezark-publish/agentdesk-mcp) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

AgentDesk MCP — Adversarial AI Review

Quality control for AI pipelines — one MCP tool. Works with Claude Code, Claude Desktop, and any MCP client.

29.5% of teams do NO evaluation of AI outputs. (LangChain Survey) Knowledge workers spend 4.3 hours/week fact-checking AI outputs. (Microsoft 2025)

AgentDesk MCP fixes this. Add independent adversarial review to any AI pipeline in 30 seconds.

Quick Start

npm (recommended)

npx @ezark-publish/agentdesk-mcp

Claude Code

claude mcp add agentdesk-mcp -- npx @ezark-publish/agentdesk-mcp

Claude Desktop

{
  "mcpServers": {
    "agentdesk-mcp": {
      "command": "npx",
      "args": ["-y", "@ezark-publish/agentdesk-mcp"],
      "env": { "ANTHROPIC_API_KEY": "sk-ant-..." }
    }
  }
}

HTTP Transport (Streamable HTTP)

Run as an HTTP server for remote access, Smithery hosting, or multi-client setups:

# Start with HTTP transport on port 3100
MCP_HTTP_PORT=3100 npx @ezark-publish/agentdesk-mcp

# Or use the --http flag (defaults to port 3100)
npx @ezark-publish/agentdesk-mcp --http

MCP endpoint: POST http://localhost:3100/mcp Health check: GET http://localhost:3100/health

Install from GitHub (alternative)

npm install github:Rih0z/agentdesk-mcp

Requirements

ANTHROPIC_API_KEY environment variable (uses your own key — BYOK)

Tools

`review_output`

Adversarial quality review of any AI-generated output. An independent reviewer assumes the author made mistakes and actively looks for problems.

Input:

Parameter	Required	Description
`output`	Yes	The AI-generated output to review
`criteria`	No	Custom review criteria
`review_type`	No	Category: `code`, `content`, `factual`, `translation`, etc.
`model`	No	Reviewer model (default: `claude-sonnet-4-6`)

Output:

{
  "verdict": "PASS | FAIL | CONDITIONAL_PASS",
  "score": 82,
  "issues": [
    {
      "severity": "high",
      "category": "accuracy",
      "description": "Claim about X is unsupported",
      "suggestion": "Add citation or remove claim"
    }
  ],
  "checklist": [
    {
      "item": "Factual accuracy",
      "status": "pass",
      "evidence": "All statistics match cited sources"
    }
  ],
  "summary": "Overall assessment...",
  "reviewer_model": "claude-sonnet-4-6"
}

`review_dual`

Dual adversarial review — two independent reviewers assess the output from different angles, then a merge agent combines findings.

If either reviewer finds a critical issue → merged verdict is FAIL
Takes the lower score
Combines and deduplicates all issues

Use for high-stakes outputs where quality is critical.

Same parameters as review_output.

How It Works

Adversarial prompting: The reviewer is instructed to assume mistakes were made. No benefit of the doubt.
Evidence-based checklist: Every PASS item requires specific evidence. Items without evidence are automatically downgraded to FAIL.
Anti-gaming validation: If >30% of checklist items lack evidence, the entire review is forced to FAIL with a capped score of 50.
Structured output: Verdict + numeric score + categorized issues + checklist (not just "looks good").

Use Cases

Code review: Check for bugs, security issues, performance problems
Content review: Verify accuracy, readability, SEO, audience fit
Factual verification: Validate claims in AI-generated text
Translation quality: Check accuracy and naturalness
Data extraction: Verify completeness and correctness
Any AI output: Summaries, reports, proposals, emails, etc.

Why Not Just Ask the Same AI to Review?

Self-review has systematic leniency bias. An LLM reviewing its own output shares the same blind spots that created the errors. Research shows models are 34% more likely to use confident language when hallucinating.

AgentDesk uses a separate reviewer invocation with adversarial prompting — fundamentally different from self-review.

Comparison

Feature	AgentDesk MCP	Manual prompt	Braintrust	DeepEval
One-tool setup	Yes	No	No	No
Adversarial review	Yes	DIY	No	No
Dual reviewer	Yes	DIY	No	No
Anti-gaming validation	Yes	No	No	No
No SDK required	Yes	Yes	No	No
MCP native	Yes	No	No	No

Limitations

Prompt injection: Like all LLM-as-judge systems, adversarial inputs could attempt to manipulate reviewer verdicts. The anti-gaming validation layer mitigates superficial gaming, but determined adversarial inputs remain a challenge. For high-stakes use cases, combine with deterministic validation.
BYOK cost: Each review_output call makes 1 LLM API call; review_dual makes 3. Factor this into your pipeline costs.

Hosted API (Separate Product)

For teams that prefer HTTP integration, a hosted REST API with additional features (agent marketplace, context learning, workflows) is available at agentdesk.usedevtools.com.

Development

git clone https://github.com/Rih0z/agentdesk-mcp.git
cd agentdesk-mcp
npm install
npm test        # 35 tests
npm run build

License

MIT

Built by EZARK Consulting | Web Version