Works with the AI Agents you already use
bd Mcp Generate Skill
Automate the secure analysis, implementation, and deployment of MCP servers for TypeScript and Next.js apps.
Free
bd-mcp-generate-skill
Example session with this skill installed
Use $mcpify in analyze mode on this Next.js repository to identify potential tools for a customer support agent.
- Read your context and instructions
- Compiled the bd-mcp-generate-
I have analyzed the repository architecture and identified key service boundaries. The analysis focused on the lib/services directory to find idempotent read operations and authorized write actions suitable for MCP exposure.
MCP Analysis Report: Customer Support Interface
| Capability | Mapping | Risk Level | Logic Source | Status |
|---|---|---|---|---|
| Get Order Status | Tool | Low | services/orders.ts:findOrder | pending_review |
| Process Refund | Tool | High | services/payments.ts:issueRefund | disabled_pending_review |
| Customer History | Resource | Low | services/customers.ts:getLogs | pending_review |
| Update Address | Tool | Medium | services/customers.ts:updateMeta | excluded |
Key Findings
Architecture: Identified Next.js App Router structure. Domain logic is well-isolated in the services layer, allowing for clean MCP wrappers.
2.
Security: The issueRefund operation requires multi-factor authorization in the base application. The generated MCP server must preserve this requirement.
3.
Exclusions: The updateMeta function was excluded as it allows writing to arbitrary JSON fields, which violates the narrow-interface invariant.
Next steps
- Review the generated
mcp-manifest.yamlin the root directory. - Change the status of desired capabilities from
pending_reviewtoapproved. - Run
$mcpify generateto implement the approved server code.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual MCP integration is prone to security overreach and architectural mess. This skill automates the full Model Context Protocol (MCP) lifecycle, transforming existing TypeScript and Next.js applications into secure, production-ready MCP servers without exposing internal logic or raw database access.
What it does
- Capability analysis Scans repositories to map business logic to MCP tools, resources, and prompts while identifying security risks.
- Server generation Implements a narrow MCP interface using existing domain services, ensuring logic isn't duplicated.
- Security auditing Inspects existing MCP manifests to flag unsafe function exposure or missing authorization boundaries.
- Automated testing Validates MCP integrations safely without using production credentials or performing destructive actions.
- Release management Handles packaging, containerization, and registry publication for platforms like the official MCP Registry.
- Operations support Manages versioning, deprecation, and security incident responses for live MCP servers.
How it works
- Analyze mode Run the skill on a repo to generate a
mcp-manifest.yamland a risk assessment report. - Approval phase The user reviews the manifest and marks specific capabilities as
approvedfor implementation. - Implementation Invoke the skill in
generatemode to write the TypeScript/Node.js server code based on the manifest. - Lifecycle execution Use
test,package,deploy, orpublishcommands to move the server toward production.
Frameworks & tools
Primarily optimized for TypeScript and Next.js repositories. Uses mcp-manifest.yaml for configuration and includes Python-based scripts for repository probing and manifest validation.
Why this beats prompting it yourself
Prompting often leads to agents exposing raw SQL or internal ORM methods as tools, creating massive security holes. This skill enforces strict invariants, requiring explicit approval for every capability and ensuring all wrappers maintain existing transaction and identity boundaries.
Use cases
- Converting a Next.js SaaS backend into a set of tools for AI agents.
- Auditing a third-party MCP server for over-permissioned access.
- Deploying a validated MCP server to a staging environment with full health checks.
- Managing the deprecation and migration path for a public MCP server listing.
Known limitations
Does not support bulk-exposure of arbitrary functions. Requires manual approval in the manifest file before code generation. Deployment and publication require explicit user authorization for each target.
Requires manual approval of entries in mcp-manifest.yaml before generation. Primarily optimized for TypeScript and Next.js; other frameworks may require manual architectural mapping.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account
Requires
Permissions
Allowed Hosts