Human Creative AI Workflow & Eval Lab

Working Security A

io.github.jdurey/human-creativev1.0.0 GitHub verified

Remote MCP tools for 3-vendor AI workflow diagnostics, HAR sandboxes, 50v audits, and rubrics.

AI & Agents Security

Watch Source Website

Registry entry

Open in the registry

{
  "_meta": {
    "io.modelcontextprotocol.registry/official": {
      "status": "active",
      "isLatest": true,
      "updatedAt": "2026-11-14T10:12:02.521076Z",
      "publishedAt": "2026-11-14T10:12:02.521076Z",
      "statusChangedAt": "2026-11-14T10:12:02.521076Z"
    }
  },
  "server": {
    "name": "io.github.jdurey/human-creative",
    "title": "Human Creative AI Workflow & Eval Lab",
    "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
    "remotes": [
      {
        "url": "https://josh-employer-funnel.josh-durey.workers.dev/mcp",
        "type": "streamable-http"
      }
    ],
    "version": "1.0.0",
    "repository": {
      "url": "https://github.com/jdurey/human-creative-mcp",
      "source": "github"
    },
    "websiteUrl": "https://jdurey.github.io",
    "description": "Remote MCP tools for 3-vendor AI workflow diagnostics, HAR sandboxes, 50v audits, and rubrics."
  }
}

Maintain this server? Claim it

Try it

Call this server's tools from your browser. Nothing to install.

https://josh-employer-funnel.josh-durey.workers.dev/mcp

Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.

Tool list from our last check. Connect live to refresh it.

audit_workflow_guardrails_50v

Runs a 50-vector multi-family prompt-injection, policy-bypass, and stem-withheld answer-leak stress test judged by 3 independent LLM vendor families, producing a hardened v2.0 system prompt and 300-DPI Executive Audit PDF ($199 Stripe / $2.50 x402 USDC, or set preview_mode=true for a free 5-vector Edge dry-run).

client_namestring, required

Organization or product name.

bot_or_workflow_namestring, required

Target assistant, agent, or workflow name.

system_prompt_or_rulesstring, required

Current system prompt, policy boundaries, or protected decision rules.

preview_modeboolean

Set true for an instant free 5-vector Edge guardrail preview.

contact_emailstring

Delivery email for the 50-vector PDF & JSON telemetry pack.

payment_tokenstring

Optional Stripe or x402 payment token.

This tool isn't marked read-only, so it may change something on the server or in an account you connected.

I understand, run it anyway

Install

Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.

One click: Add to Cursor Add to VS Code Add to LM Studio Add to Goose

Run in a terminal

claude mcp add --transport http human-creative https://josh-employer-funnel.josh-durey.workers.dev/mcp

Tools (5)

audit_workflow_guardrails_50v

Argument	Type	Description
client_name*	string	Organization or product name.
preview_mode	boolean	Set true for an instant free 5-vector Edge guardrail preview.
contact_email	string	Delivery email for the 50-vector PDF & JSON telemetry pack.
payment_token	string	Optional Stripe or x402 payment token.
bot_or_workflow_name*	string	Target assistant, agent, or workflow name.
system_prompt_or_rules*	string	Current system prompt, policy boundaries, or protected decision rules.

Input schema (JSON)

{
  "type": "object",
  "required": [
    "client_name",
    "bot_or_workflow_name",
    "system_prompt_or_rules"
  ],
  "properties": {
    "client_name": {
      "type": "string",
      "description": "Organization or product name."
    },
    "preview_mode": {
      "type": "boolean",
      "default": false,
      "description": "Set true for an instant free 5-vector Edge guardrail preview."
    },
    "contact_email": {
      "type": "string",
      "description": "Delivery email for the 50-vector PDF & JSON telemetry pack."
    },
    "payment_token": {
      "type": "string",
      "description": "Optional Stripe or x402 payment token."
    },
    "bot_or_workflow_name": {
      "type": "string",
      "description": "Target assistant, agent, or workflow name."
    },
    "system_prompt_or_rules": {
      "type": "string",
      "description": "Current system prompt, policy boundaries, or protected decision rules."
    }
  }
}

build_rea_har_agent_sandbox

Converts a live web workflow URL or recorded HTTP Archive (.har v1.2) into an offline deterministic REA AgentEnv MCP sandbox (callable read/write tools + state transitions) paired with 8 multi-step evaluation tasks, 3-vendor Judge-Trust reward verifiers, a 300-DPI Executive Benchmark PDF, and a SHA-256 verified .zip bundle ($299 Stripe / $3.50 x402 USDC, or set preview_mode=true for an instant free Edge dry-run).

Argument	Type	Description
target_url	string	Optional HTTPS workflow URL to capture into a validated HAR v1.2 archive.
client_name*	string	Organization or team name commissioning the AgentEnv sandbox.
preview_mode	boolean	Set true to run a free instant Edge dry-run returning projected MCP tools and 8-task benchmark outline.
contact_email	string	Delivery email for the compiled AgentEnv .zip and 300-DPI Executive PDF.
payment_token	string	Optional Stripe checkout session / payment_intent ID or x402 USDC payment receipt.
workflow_name*	string	Name of the multi-step web or API workflow (e.g. 'Gift Cart & Promo Checkout').
observed_endpoints	array	Optional list of key REST/JSON endpoints (e.g. 'GET /v1/gifts', 'POST /v1/orders').

Input schema (JSON)

{
  "type": "object",
  "required": [
    "client_name",
    "workflow_name"
  ],
  "properties": {
    "target_url": {
      "type": "string",
      "description": "Optional HTTPS workflow URL to capture into a validated HAR v1.2 archive."
    },
    "client_name": {
      "type": "string",
      "description": "Organization or team name commissioning the AgentEnv sandbox."
    },
    "preview_mode": {
      "type": "boolean",
      "default": false,
      "description": "Set true to run a free instant Edge dry-run returning projected MCP tools and 8-task benchmark outline."
    },
    "contact_email": {
      "type": "string",
      "description": "Delivery email for the compiled AgentEnv .zip and 300-DPI Executive PDF."
    },
    "payment_token": {
      "type": "string",
      "description": "Optional Stripe checkout session / payment_intent ID or x402 USDC payment receipt."
    },
    "workflow_name": {
      "type": "string",
      "description": "Name of the multi-step web or API workflow (e.g. 'Gift Cart & Promo Checkout')."
    },
    "observed_endpoints": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Optional list of key REST/JSON endpoints (e.g. 'GET /v1/gifts', 'POST /v1/orders')."
    }
  }
}

calibrate_judge_rubric_4family

Eliminates single-model self-preference bias and stochastic verdict flips by decomposing subjective quality criteria into deterministic Python checks + a 4-family Judge-Trust rubric panel verified across byte-identical repeats ($149 Stripe / $2.00 x402 USDC, or set preview_mode=true for a free calibrated rubric preview).

Argument	Type	Description
client_name*	string	Organization or team name.
task_domain*	string	Output type being graded (e.g. 'Customer refund drafts', 'SQL agent traces').
preview_mode	boolean	Set true for an instant free binary pass/fail rubric decomposition on Edge.
contact_email	string	Delivery email for the calibrated rubric & repeatability report.
payment_token	string	Optional Stripe or x402 payment token.
current_rubric*	string	Current grading prompt, Likert scale, or QA checklist.

Input schema (JSON)

{
  "type": "object",
  "required": [
    "client_name",
    "task_domain",
    "current_rubric"
  ],
  "properties": {
    "client_name": {
      "type": "string",
      "description": "Organization or team name."
    },
    "task_domain": {
      "type": "string",
      "description": "Output type being graded (e.g. 'Customer refund drafts', 'SQL agent traces')."
    },
    "preview_mode": {
      "type": "boolean",
      "default": false,
      "description": "Set true for an instant free binary pass/fail rubric decomposition on Edge."
    },
    "contact_email": {
      "type": "string",
      "description": "Delivery email for the calibrated rubric & repeatability report."
    },
    "payment_token": {
      "type": "string",
      "description": "Optional Stripe or x402 payment token."
    },
    "current_rubric": {
      "type": "string",
      "description": "Current grading prompt, Likert scale, or QA checklist."
    }
  }
}

compile_sop_to_scorm_qti

Compiles raw Markdown/text SOPs, runbooks, or training docs into LMS-ready SCORM 1.2 + 1EdTech QTI 3.0 XML mastery packages with stem-withheld blind-QA item verification ($49 Stripe / $0.75 x402 USDC, or set preview_mode=true for a free Knowledge Component + QTI 3.0 XML item preview).

Argument	Type	Description
client_name	string	Organization or training team name.
course_title*	string	Title of the SOP module or mastery course.
preview_mode	boolean	Set true for an instant free Knowledge Component + QTI 3.0 XML item preview.
sop_markdown*	string	Source SOP, policy, or instructional Markdown text.
contact_email	string	Delivery email for validated SCORM 1.2 and QTI 3.0 .zip archives.
payment_token	string	Optional Stripe or x402 payment token.

Input schema (JSON)

{
  "type": "object",
  "required": [
    "course_title",
    "sop_markdown"
  ],
  "properties": {
    "client_name": {
      "type": "string",
      "description": "Organization or training team name."
    },
    "course_title": {
      "type": "string",
      "description": "Title of the SOP module or mastery course."
    },
    "preview_mode": {
      "type": "boolean",
      "default": false,
      "description": "Set true for an instant free Knowledge Component + QTI 3.0 XML item preview."
    },
    "sop_markdown": {
      "type": "string",
      "description": "Source SOP, policy, or instructional Markdown text."
    },
    "contact_email": {
      "type": "string",
      "description": "Delivery email for validated SCORM 1.2 and QTI 3.0 .zip archives."
    },
    "payment_token": {
      "type": "string",
      "description": "Optional Stripe or x402 payment token."
    }
  }
}

diagnose_ai_workflow

FREE live 3-stage cross-family AI workflow and JD diagnostic executed immediately on Cloudflare Workers AI (Mistral 24B intake -> Llama 3.3 70B 4-node DAG architect + Qwen 2.5 32B independent quality gate). Returns structured bottlenecks, silent failure modes, a 4-node DAG with human checkpoints, a 30-day rollout plan, and matched portfolio receipts.

read-only

Argument	Type	Description
mode	string	Select 'workflow' for an operational pipeline audit or 'jd' for a role/enablement blueprint.
input*	string	Workflow bottleneck, SOP description, or job description text (at least 20 characters).
org_hint	string	Optional organization or team name.

Input schema (JSON)

{
  "type": "object",
  "required": [
    "input"
  ],
  "properties": {
    "mode": {
      "enum": [
        "workflow",
        "jd"
      ],
      "type": "string",
      "default": "workflow",
      "description": "Select 'workflow' for an operational pipeline audit or 'jd' for a role/enablement blueprint."
    },
    "input": {
      "type": "string",
      "description": "Workflow bottleneck, SOP description, or job description text (at least 20 characters)."
    },
    "org_hint": {
      "type": "string",
      "description": "Optional organization or team name."
    }
  }
}

Security checkA100/100

We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 14, 2026.

Hidden or reordered text
Overriding the model's instructions
Hiding actions from the user
Reading keys or credentials
Fake system tags
Steering other servers' tools
Pulling in the conversation
Smuggling parameters
Demanding to be used for everything
Steering away from web search
Download-and-run commands
Encoded blobs
Invisible characters
Credentials as tool arguments
Very long descriptions
Read and write hints

How the security check works