Spedy NFS-e MCP
Enables issuing, consulting, and listing Brazilian NFS-e invoices through Spedy's API over an authenticated HTTP endpoint.
README
Spedy NFS-e MCP
An Invokta Action Engine that exposes Spedy NFS-e emission and status operations through an authenticated MCP Streamable HTTP endpoint.
The engine contains three capabilities:
openapi.emitir-nfse: creates and enqueues an NFS-e for asynchronous issue.openapi.consultar-nfse: reads the current status of one NFS-e.openapi.listar-nfse: lists NFS-e records with Spedy filters and pagination.
All capabilities require an authenticated principal. Spedy credentials remain server-side and are never accepted as MCP tool input.
First run
npm run check
npm run build
npm run mcp:http
The included local configuration serves MCP at http://127.0.0.1:3400/mcp. Clients
must send Authorization: Bearer <MCP_BEARER_TOKEN>.
The committed .env.example and the ignored local .env contain mock values.
Replace MCP_BEARER_TOKEN and SPEDY_API_KEY before making real requests. The
default connector target is Spedy's sandbox; production is never selected
implicitly.
Configuration
| Variable | Purpose |
|---|---|
MCP_BEARER_TOKEN |
Protects the MCP HTTP endpoint; minimum 32 characters. |
SPEDY_API_KEY |
Sent by the connector as the server-side X-Api-Key. |
SPEDY_BASE_URL |
Spedy API origin, normally https://sandbox-api.spedy.com.br. Do not append /v1. |
INVOKTA_HTTP_HOST |
Bind host; defaults to 127.0.0.1. |
INVOKTA_HTTP_PORT |
Bind port; the local example uses 3400 and the adapter defaults to 3000 when unset. |
Emission is asynchronous. A successful emitir-nfse result means Spedy
accepted and enqueued the request; use consultar-nfse or Spedy webhooks to
observe authorization or rejection. Always provide a stable integrationId
to make retries idempotent.
Project map
src/capabilities/owns domain actions and their input, output, access, and execution contracts.src/engine.tsregisters stable capability IDs and supplies engine-owned dependencies.src/direct.tsis the smallest executable smoke path.test/engine.test.tsproves behavior through the public engine boundary.AGENTS.mdrecords the project's non-negotiable architecture and delivery rules..agents/skills/develop-invokta-project/SKILL.mdgives compatible agents a complete change workflow.
Review the generated capabilities
- Read
AGENTS.mdand the generated development skill. - Add a failing engine-level test for the domain outcome you need.
- Update or replace the module under
src/capabilities/, then register its literal capability ID insrc/engine.ts. - Keep direct and generated adapter entry points on the same exported engine; do not duplicate the capability handler in an adapter.
- Run
npm run check, then inspect the capability in the devtools before deploying it.
Develop interactively
npm run devtools
Builds the engine and starts the Invokta devtools, normally on http://localhost:4100/. Open the URL printed by the command to browse capabilities, invoke them from schema-seeded JSON through the execution path you select, follow the live invocation trace, switch test identities, and read the doctor report. Source changes rebuild and restart the hosted engine automatically.
Run the read-only devtools diagnostics with:
npm run devtools:doctor
Start MCP HTTP
npm run build
npm run mcp:http
Package the server and probe an existing endpoint with:
npm run deploy:package
npm run deploy:probe -- --url https://engine.example/mcp
Review every generated capability's domain meaning and access rule before deployment.
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.