K-Work Trust Public Demo
Remote MCP/API trust gate for AI agents to inspect readiness before performing Korean business workflow operations like vendor onboarding and payment release.
README
K-Work Trust Public Demo
K-Work Trust is a remote MCP/API trust gate for AI agents before they write Korean business workflow records.
It helps an external AI agent inspect whether it should continue, pause, retry, stop, or escalate before registering vendors, checking invoice readiness, releasing payment or invoice steps, updating CRM, or touching ecommerce back-office records.
Public Demo
- Demo page: https://k-work-trust-api.fly.dev/agent-demo
- Remote MCP endpoint: https://k-work-trust-api.fly.dev/mcp
- Readiness: https://k-work-trust-api.fly.dev/v1/agent/readiness
- Agent manifest: https://k-work-trust-api.fly.dev/.well-known/k-work-trust-agent.json
- Skill instructions: https://k-work-trust-api.fly.dev/skill.md
- OpenAPI: https://k-work-trust-api.fly.dev/openapi.json
What To Test
- Read the public readiness endpoint.
- Read the agent manifest, skill instructions, and OpenAPI.
- Use
POST https://k-work-trust-api.fly.dev/mcpif your client supports MCP Streamable HTTP. - Use provisional signup if your agent has no inbox, or email-verified signup if it can read email.
- Call
/v1/auth/whoamibefore spending quota. - Run one synthetic
/v1/vendor-onboarding/validatecall. - Report only non-secret debugging fields.
Safe Reporting Fields
request_id- endpoint called
- HTTP status
- response
statusanddecision - first blocked step
- confusing or missing response field
Do not post API keys, proof solutions, email verification codes, private customer records, production screenshots, or sensitive business data.
Synthetic Payload
See examples/synthetic-vendor.json. It is fake test data and is safe to copy into a local agent experiment.
Minimal Metadata Fetch
node examples/fetch-public-metadata.js
The script reads public metadata only. It does not require an API key.
Remote MCP
Remote MCP supports initialize, tools/list, and tools/call over JSON-response Streamable HTTP. initialize and tools/list are public for discovery; MCP tools/call uses the same API key, quota, official-source, and abuse-guard policies as protected REST workflow calls.
Product Gates
validate_vendor_onboarding: decide whether a Korean vendor/customer record can be registered.validate_payment_or_invoice_release: decide whether an agent can release a payment, invoice, settlement, or purchase step, or whether it must prepare a human review packet first.
Why This Exists
Mature KYB platforms are broad compliance systems for onboarding, UBO, sanctions, adverse media, and case review. K-Work Trust is narrower: a small agent-readable pre-action gate with stable JSON decisions, public MCP discovery, low-quota test keys, and Korean business-field logic.
The core response contract is meant for another agent to use directly:
decisionblocking_issueswarningsevidenceagent_next_action
Scope
This public demo repo intentionally excludes private operating memory, API usage logs, credentials, .env, local model files, and customer data. The production service enforces authentication, low beta quotas, and behavior-based abuse blocking for protected workflow calls.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.