AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
TypeScript ai-agentai-agent-toolsai-agentsai-agents-frameworkmcpmcp-servermcp-toolsmcpsn8n-alternativeno-code-automationworkflowworkflow-automationworkflows
import { Readable } from 'node:stream'
import { buffer as streamToBuffer } from 'node:stream/consumers'
import { XMLParser } from 'fast-xml-parser'
import { FastifyPluginAsync, FastifyRequest } from 'fastify'
import { webhookController } from './webhook-controller'
import { isBinaryContentType, isMultipartContentType } from './webhook-request-converter'
export const webhookModule: FastifyPluginAsync = async (app) => {
// Capture rawBody (for signature verification) only for the small, string-parsed content
// types. Streamed types (multipart, binary files) are consumed straight to storage and get
// no rawBody — this replaces fastify-raw-body, which buffered the whole body and defeated streaming.
app.addHook('preParsing', async (request, _reply, payload) => {
// isMultipart() isn't set this early (it's flagged during content-type parsing), so
// detect streamed types from the header directly and leave their stream untouched.
const contentType = request.headers['content-type']
if (isMultipartContentType(contentType) || isBinaryContentType(contentType)) {
return payload
}
const raw = await streamToBuffer(payload)
request.rawBody = raw.toString('utf8')
return Readable.from(raw)
})
app.addContentTypeParser(
'application/json',
{ parseAs: 'string' },
(_req, body: string, done) => {
if (body == null || body.trim() === '') {
return done(null, {})
}
try {
done(null, JSON.parse(body))
}
catch (err) {
const error: Error & { statusCode?: number } = err instanceof Error ? err : new Error('JSON parsing failed')
error.statusCode = 400
done(error, undefined)
}
},
)
// Binary bodies pass through as a raw stream so webhook-request-converter can stream them
// straight to storage instead of buffering the whole file in memory.
const streamThroughBinary = (_request: FastifyRequest, payload: Readable, done: (err: Error | null, body?: Readable) => void) => done(null, payload)
app.addContentTypeParser(/^image\/.*/, streamThroughBinary)
app.addContentTypeParser(/^video\/.*/, streamThroughBinary)
app.addContentTypeParser(/^audio\/.*/, streamThroughBinary)
app.addContentTypeParser('application/pdf', streamThroughBinary)
app.addContentTypeParser('application/zip', streamThroughBinary)
app.addContentTypeParser('application/gzip', streamThroughBinary)
app.addContentTypeParser('text/csv', streamThroughBinary)
// octet-stream has a global buffering parser; override it (scoped to webhook routes) to stream.
app.removeContentTypeParser('application/octet-stream')
app.addContentTypeParser('application/octet-stream', streamThroughBinary)
// processEntities: false prevents DOCTYPE entity declarations from overriding
// built-in XML entities (< > & etc), blocking injection attacks.
const xmlParser = new XMLParser({ processEntities: false })
app.addContentTypeParser(
['text/xml', 'application/xml', 'application/rss+xml'],
{ parseAs: 'string' },
(_req, body: string, done) => {
try {
done(null, xmlParser.parse(body))
}
catch (err) {
const error: Error & { statusCode?: number } = err instanceof Error ? err : new Error('XML parsing failed')
error.statusCode = 400
done(error, undefined)
}
},
)
await app.register(webhookController, { prefix: '/v1/webhooks' })
}