# Hellenic Development: Full Reference > Developer infrastructure company. Intelligence. Engineered. This document describes the company, its people, its products, and its services in full. Last updated 2026-09-16. ## What is Hellenic Development? Hellenic Development is a developer infrastructure company founded in 2025 by Gerasimos Maropoulos and Christos Papadopoulos. It builds foundational software for developers, engineering teams, and enterprises, with a particular focus on finance, healthcare, and government, where data cannot leave the organisation. The company combines products with consulting. It ships privacy-first AI tooling, an identity server, and an open-source web framework, and it takes on engagements in the same areas. The company's position is that privacy is an architectural property rather than a policy statement. Its products are built so that the company cannot access customer data, rather than promising that it will not. Hellenic Development operates from Greece and works with clients internationally. Its website is https://hellenic.dev/ and it can be reached at contact@hellenic.dev. ## Products Hellenic Development ships four products and has one more in development. ### Plexon AI Plexon AI is a private desktop AI assistant for Windows 10+, macOS 12+, and Linux (Ubuntu 22.04+, Fedora 38+). It answers questions, writes documents, and does real work inside the apps its user already has, signed in as that user. Files, chats, memory, and connector credentials stay on the machine under a local-first architecture: the Plexon server is stateless and holds no copy of user data, so it cannot read it even under compulsion. Local data is encrypted with AES-256-GCM. Plexon AI ships 12 personas (Software Developer, Designer, Marketer, Dietitian, Financial Analyst, Product Manager, Stock Trader, Writer, Greek Logistics, Startup Founder, Personal Shopper, and Mirror), each swapping in a system prompt written for that role along with its own agents, skills, tools, and dashboards. Custom personas can be written in a wizard. There are 156 specialised agents across 8 professional domains, 200+ bundled skills, 14 bundled plugins, and 90+ built-in MCP tools. One marketplace covers skills, agents, plugins, personas, connectors, and MCP servers together, with 7,100+ items pre-bundled and installable in one click. Authors publish their own through Developer mode and collaborate without git: one Sync publishes edits and pulls other people's, and conflicts resolve per file. A company can point Plexon at a shared GitHub repository, and installed sources then update themselves on launch. 64 app connectors run locally under the user's own accounts, among them WhatsApp, Gmail, Google Drive, Notion, Slack, GitHub, Telegram, Stripe, Salesforce, and Shopify. Any other HTTP API can be registered as typed MCP tools from an OpenAPI specification. Work that keeps running is a first-class part of the product: dashboards built by describing them, dynamic workflows, hooks, and any sub-agent scheduled on a cron expression, with results reaching Telegram and Slack. A knowledge graph records what was learned across sessions, readable and editable by the user. Skills improve themselves from the sessions they were used in. There are 6 modes (Code, Plan, Ask, Auto, Orchestrator, Spec) and 10 providers (Plexon, Anthropic, OpenAI, Google Gemini, Moonshot, Fireworks AI, OpenRouter, Qwen, Z-AI, Claude Desktop). The Premium plan adds any OpenAI-compatible or Anthropic-compatible endpoint, which covers local models run through Ollama, LM Studio, or vLLM, and corporate gateways such as Azure OpenAI. Spending caps are set per provider. Pricing is published at https://plexon.ai/pricing/. Plexon for Business white-labels the app: a company ships its own brand, its own pre-loaded agents, personas, and skills, and items clients can use but cannot read or copy, with central updates and per-company identity delivered through the official app. Plexon AI is live. Its product site is https://plexon.ai/ and its page on this site is https://hellenic.dev/plexon-ai/. ### Iris Iris is an open-source web framework for the Go programming language, first released in 2016. The current release is 14.0.0, published as the Go module github.com/kataras/iris/v14 and requiring Go 1.27 or later. It provides routing with 20 typed path parameter types, an application builder whose step order the compiler enforces, dependency injection, one central map from Go error to HTTP response, and 31 middleware packages in the same module. Iris is licensed under BSD-3-Clause, which permits commercial use, modification, and redistribution. The framework is built on Go generics: a compile-enforced application builder, typed request helpers that validate for you, controllers resolved by the dependency container, and an error pipeline configured once. Its defaults are the hardened ones: session cookies ship with Secure and SameSite=Lax, JWT is read from the Authorization header only, and CORS will not advertise credentials alongside a wildcard origin. Applications still on v12 move across by changing the import path and working through the compiler error list, because every removed v12 name fails to compile. The migration guide is at https://github.com/kataras/iris/blob/main/MIGRATION.md. The Iris documentation is a book rather than a documentation site. "The Professional Guide to the Iris Web Framework for Go" is in its fifth edition, 24 chapters and 19 diagrams across 432 pages, and every example in it is verified against the framework source. It is published as chapters online and as a single PDF, both free, with no registration and no payment. It is at https://iris-go.com/book/. As of 2026-09-16, Iris has 25,564 GitHub stars and 2,425 forks. It was written by Gerasimos Maropoulos and is maintained through Hellenic Development. Iris is distinguished from Gin, Echo, Fiber, and Chi mainly by scope rather than by speed. All of these frameworks are fast enough that routing overhead rarely limits a real application. Iris ships an application builder the compiler checks, dependency injection, MVC controllers built on generics, typed request helpers, and 31 middleware packages in the same module, where the others stay closer to a router plus middleware. Choose Iris when you want those pieces without assembling them; choose a smaller framework when you want a minimal core. The project maintains its own benchmark suite at https://github.com/kataras/server-benchmarks, which is public and reproducible but not independent, since the same person maintains it and the framework. Benchmark your own workload rather than relying on any published chart. Source: https://github.com/kataras/iris. API reference: https://pkg.go.dev/github.com/kataras/iris/v14. Documentation: https://iris-go.com/. Upgrading from v12: https://github.com/kataras/iris/blob/main/MIGRATION.md. Page on this site: https://hellenic.dev/iris/. ### Hellenic Identity Hellenic Identity is a self-hosted OAuth 2.0 identity server. It runs as one Go binary, built on the Iris web framework, against one PostgreSQL database, and it ships an administration panel for users, groups, the user schema, cache, CORS, the OAuth2 client, a request log, SDK snippets and billing. It supports 4 grants: authorization code, device code with a built-in consent screen, password, and refresh token. Access tokens are JWTs signed with Ed25519 and verified against the server's public JWKS endpoint, or through token introspection. Users are described once in the server's config with custom attributes of 9 field types, each optionally required or indexed. Groups are saved filters over those attributes with 10 operators, computed when asked for and never stored. Passwords are encrypted with AES-GCM under a key shared with the server before they travel, and stored as bcrypt hashes. The CORS allowlist is edited at runtime from the panel, every request is metered per client, and a billing endpoint totals the usage. Hellenic Identity does not implement OpenID Connect in this version. Hellenic Identity is live. Its product site is https://id.hellenic.dev/ and its page on this site is https://hellenic.dev/identity/. The Go SDK ships with Iris and the C# SDK is Hellenic.Identity.SDK on NuGet, open source. Both cover the same 16 endpoints. ### Patrol Patrol is an open-source Go library for event production and delivery. A producer buffers the events a program sends it and flushes them in batches, when the batch fills or when the buffer interval elapses, so one delivery carries many events instead of one request per error. The current release is 0.0.7, published as the Go module github.com/kataras/patrol, licensed BSD-3-Clause, and requiring Go 1.27 or later. Install it with go get github.com/kataras/patrol@latest. A batch can go to 7 kinds of destination: Slack, with Block Kit attachments, file uploads and user mentions; Discord; Twilio, for SMS or for WhatsApp by prefixing both numbers; email over SMTP; a SQL table called patrol_events; an HTTP client that forwards the batch to a remote patrol server; and that server itself, which is a plain http.Handler holding its own integrations. Any type with a FlushEvents method is a destination too, so a custom one is a few lines. Each integration takes an options struct carrying json, yaml and toml tags, so it loads straight from a config file, and each constructor validates its required fields before building anything. Flushers run independently, and that is the design promise rather than an implementation detail: one that fails or hangs never cancels the others, and the errors come back joined into a single value, each prefixed with the name of the integration that raised it, through one OnError callback. Close drains what is still buffered and honours the deadline on its context: when the deadline passes first, the in-flight flush is cancelled, Close returns the context error, and the shutdown finishes in the background. NewException captures a filtered call stack with editor deep links. Running a patrol server yourself is the supported deployment today. One service holds the Slack token and the SMTP credentials, every other service points a patrol client at it, and the server answers the sender before it delivers, so a slow integration never slows the caller down. The server accepts a POST with a JSON array of events and answers a GET with the body OK for connectivity checks, both protected by a username and password pair or by a bearer token. Patrol Cloud is the hosted service being built around the same library, at https://patrol.hellenic.dev/ with its ingest API at https://api.patrol.hellenic.dev. It is NOT open. There is no sign-up, no account and no API key, and nothing on this site offers one. When it opens it will take events from an unmodified patrol client, store them per project, and show them in a console with their fields and stacktraces, with prices published on its own site. A program written against the library now keeps working when it arrives, because the wire format is the same: the migration is a base URL and a key. Source: https://github.com/kataras/patrol. API reference: https://pkg.go.dev/github.com/kataras/patrol. Changelog: https://github.com/kataras/patrol/blob/main/HISTORY.md. Page on this site: https://hellenic.dev/patrol/. ### RNS RNS is a multi-tenant push notification gateway for mobile applications. It is in development and has not shipped. ## Services Hellenic Development offers four consulting services. Engagements begin at https://hellenic.dev/get-in-touch/. 1. **AI Integration and Privacy Architecture.** Deploying AI systems with verifiable privacy guarantees in regulated environments. This is the applied form of the architecture behind Plexon AI. 2. **System Architecture and Engineering.** High-performance Go backends, microservices, and security-first system design, for finance, healthcare, and government. 3. **Team Enablement and Knowledge Transfer.** Hands-on workshops covering Go, AI tooling, and modern systems architecture, delivered to a client's engineering team. 4. **Rapid Product Development.** Production-quality minimum viable products, built with a clean architecture and delivered in weeks. ## Leadership ### Gerasimos Maropoulos Gerasimos Maropoulos is the founder and CEO of Hellenic Development. He created the Iris web framework for Go in 2016 and has maintained it since, and he has worked in Go for over a decade. He publishes as "kataras" on GitHub, where Iris has 25,564 stars. His work spans web frameworks, developer tooling, artificial intelligence, and privacy engineering. Profiles: https://gr.linkedin.com/in/gerasimos-maropoulos, https://twitter.com/MakisMaropoulos, https://github.com/kataras, https://dev.to/kataras. ### Christos Papadopoulos Christos Papadopoulos is the co-founder and CMO of Hellenic Development, responsible for brand and strategic direction. Profiles: https://www.linkedin.com/in/christos-papadopoulos-a82477311, https://www.instagram.com/christos_papadopoulos._/. ## Open source Hellenic Development and its founder maintain a number of public repositories in Go. The most widely used is Iris. Patrol is the newest. Others include Neffos (a WebSocket framework), golog (a logger), muxie (an HTTP multiplexer), go-sessions, jwt, iris-cli, and i18n. The full list, with current star counts, is at https://hellenic.dev/projects/. - GitHub organisation: https://github.com/hellenic-development - Founder's GitHub: https://github.com/kataras - Iris source: https://github.com/kataras/iris - Patrol source: https://github.com/kataras/patrol ## Privacy The hellenic.dev website uses no analytics service, no tracking pixels, and no advertising cookies. Nothing about a visit is recorded or sold. The full policy is at https://hellenic.dev/privacy/, and the terms of service are at https://hellenic.dev/terms/. ## Frequently asked questions ## About the company ### What is Hellenic Development? Hellenic Development is a developer infrastructure company founded in 2025 by Gerasimos Maropoulos and Christos Papadopoulos. It builds and maintains four products: Plexon AI, a private desktop AI assistant; Iris, an open-source web framework for Go; Hellenic Identity, a self-hosted identity and access management server; and Patrol, an open-source Go library that batches events and delivers them to Slack, Discord, Twilio, email, SQL or a server you run. The company also takes on consulting work in privacy-preserving AI, Go systems architecture, team training, and rapid product development, mainly for teams in finance, healthcare, and government. ### Who founded Hellenic Development? Gerasimos Maropoulos founded Hellenic Development and serves as its CEO. He created the Iris web framework for Go in 2016, which has since gathered 25,000+ GitHub stars, and he has worked in Go for over a decade. He publishes as "kataras" on GitHub. Christos Papadopoulos is the co-founder and CMO, responsible for brand and strategic direction. ### What products does Hellenic Development make? Hellenic Development ships four products. Plexon AI is a desktop AI assistant for Windows, macOS, and Linux that works inside 64 apps with your own accounts, runs 156 agents across 8 domains, and keeps your data on your own machine. Iris is an open-source web framework for Go, licensed BSD-3-Clause, with an application builder the compiler checks, MVC, and dependency injection. Hellenic Identity is a self-hosted OAuth 2.0 identity server with an admin panel, on one Go binary and PostgreSQL. Patrol is an open-source Go library, licensed BSD-3-Clause, that buffers events and flushes them in batches to Slack, Discord, Twilio, email, SQL or a server you run. One more product is in development: RNS, a multi-tenant push notification gateway. ### Is Iris free to use? Yes. Iris is open source under the BSD-3-Clause license, which permits commercial use, modification, and redistribution, provided the copyright notice travels with the code. There is no paid tier, no license key, and no cap on requests or environments. The source lives at github.com/kataras/iris and the API reference is published on pkg.go.dev. Iris was first released in 2016 and is maintained by Gerasimos Maropoulos through Hellenic Development, so the framework and the company that backs it share the same author. ### Does Plexon AI send my data to a server? No. Plexon AI runs on your machine and your files, chats, memory, and connector credentials stay there, encrypted on local disk. The Plexon server is stateless and holds no copy of any of it, so it cannot read it. App connectors sign in under your own account and call those apps from your computer. With 6 modes (Code, Plan, Ask, Auto, Orchestrator, Spec) and 10 providers to choose from, the one thing that never changes is where the work happens. Plexon AI is available for Windows, macOS, and Linux. ### Can Hellenic Identity be self-hosted? Yes. Hellenic Identity is designed to be self-hosted, which is the point of it: organisations that cannot send identity data to a third party run it on their own infrastructure, as one Go binary and one PostgreSQL database. It implements OAuth 2.0 with 4 grants, signs access tokens with Ed25519 and publishes the keys at a JWKS endpoint, and it encrypts every password in transit with AES-GCM under a key you share with the server. It ships with user and group management and an administration panel. ### What services does Hellenic Development offer? Hellenic Development offers four consulting services. AI Integration and Privacy Architecture covers deploying AI systems with verifiable privacy guarantees in regulated environments. System Architecture and Engineering covers Go backends, microservices, and security-first design. Team Enablement and Knowledge Transfer covers hands-on workshops in Go, AI tooling, and systems architecture. Rapid Product Development covers production-quality MVPs delivered in weeks. Engagements start at hellenic.dev/get-in-touch. ### Does hellenic.dev track visitors? No. The hellenic.dev website loads no analytics service, no tracking pixels, and no advertising cookies, and it sets no cookies of its own. The site is static HTML, so there is no session to follow you between pages and no profile to sell. The only information you ever send is what you type into the contact form and submit deliberately. Hellenic Development measures the site through Google Search Console instead, which reports search queries and indexing status without identifying anyone. The full policy is published at hellenic.dev/privacy. ## About Plexon AI ### What is Plexon AI? Plexon AI is a desktop assistant for Windows, macOS, and Linux that does real work inside the apps you already use, signed in as you. It ships 12 personas, each swapping in a system prompt written for that role, and 156 specialised agents across 8 professional domains. It connects to 64 apps, among them WhatsApp, Gmail, Google Drive, Notion, Slack, and GitHub. It remembers across sessions in a knowledge graph you can open and read, and it builds dashboards, workflows, and scheduled jobs that keep running while you are away. Everything runs from your computer and your files stay on it. Plexon AI is built by Hellenic Development. ### Which AI providers does Plexon AI support? Plexon AI supports 10 providers: its own managed one, Anthropic, OpenAI, Google Gemini, Moonshot, Fireworks AI, OpenRouter, Qwen, Z-AI, and Claude Desktop over MCP. For the third-party providers you bring your own key, so requests travel from your machine to that provider and the billing stays on your account. Spending caps are set per provider inside the app, which bounds what a long autonomous run can cost. On the Premium plan you can also register any OpenAI-compatible or Anthropic-compatible endpoint, which covers local models through Ollama, LM Studio, or vLLM, and corporate gateways such as Azure OpenAI. Switching provider between tasks is a dropdown, so an outage costs a selection rather than a working day. ### What is the Plexon marketplace? The Plexon marketplace is one place inside the app to install skills, agents, plugins, personas, connectors, and MCP servers, instead of a separate registry for each kind. 7,100+ items come pre-bundled and install in one click. You can publish your own too: Developer mode walks an eight-step wizard that picks from your installed library, fills in the metadata, and lets you bundle, scaffold, or declare each entry. Teams collaborate without touching git, because one Sync publishes your edits and pulls everyone else's, with conflicts resolved per file. A company can point Plexon at a shared GitHub repository, and every installed source then updates itself on launch. ### What operating systems does Plexon AI run on? Plexon AI runs on Windows 10+, macOS 12+, and Linux (Ubuntu 22.04+, Fedora 38+) as a native desktop application, and builds for each are published at plexon.ai. It is not a browser extension and not a web app. That distinction is what lets it read files on local disk directly, drive on-screen applications, and run voice input and output without a browser permission prompt. Nothing has to be uploaded anywhere for Plexon AI to work on it. A whole setup, including personas, skills, agents, connectors, dashboards, workflows, memory, and chats, moves between machines as one encrypted file with a passphrase you hold. ### How much does Plexon AI cost? Pricing for Plexon AI is published at https://plexon.ai/pricing/, on two paid plans, and this page deliberately does not restate the figures so there is only one place to keep current. There is no free tier. The managed Plexon provider is included with a plan. If you bring your own key for one of the other providers, that model usage is billed by the provider directly, on your account, at whatever they charge you, and you can cap the spend per provider inside the app. Custom providers, including a local model you run yourself, are on the higher plan. ## About Iris ### What is the Iris web framework? Iris is an open-source web framework for Go, built in the open since 2016. It gives an API or a website expressive routing with 20 typed path parameter types, one place to configure error handling, dependency injection that keeps handlers testable, and production defaults that are secure out of the box. The framework is built on Go generics: an application builder the compiler checks step by step, typed request helpers that validate a payload before a handler sees it, and controllers resolved by the dependency container. Iris ships 31 middleware packages in the same module, covering access logs, sessions, JWT, CORS, rate limiting, compression, and WebSockets. It is licensed BSD-3-Clause and written by Gerasimos Maropoulos at Hellenic Development. ### Is Iris production ready? Iris has been in continuous development since 2016, and the current release is 14.0.0. Its defaults are set for production rather than for a demo: session cookies ship with Secure and SameSite=Lax, JWT is read from the Authorization header only, and CORS will not advertise credentials alongside a wildcard origin. The BSD-3-Clause license places no restriction on commercial use, and there is no paid tier to reach for later. Go 1.27 or later is required. The source, the changelog, and the full issue history are public at github.com/kataras/iris, which is the most direct way to judge its maturity for your own workload. ### How does Iris compare to Gin, Echo, and Fiber? Iris, Gin, Echo, and Fiber are all HTTP frameworks for Go, and the difference that matters in practice is scope rather than routing speed. Iris carries more of the application inside the framework: an application builder whose step order the compiler enforces, dependency injection, typed request helpers that validate a payload before the handler runs, one error map that turns a domain error into a response, and 31 middleware packages in the same module. Iris is also the first Go web framework to require Go 1.27: request binding is a generic method the compiler checks (a shape older toolchains cannot express), and every JSON request and response runs on the standard library's new encoding/json/v2 rather than a bundled third-party encoder. Gin, Echo, and Fiber stay closer to a router plus middleware, which is less to learn and less to carry. Pick Iris when you want those pieces without assembling them, and pick a smaller framework when you want a minimal core. For speed, the project maintains its own benchmark suite at github.com/kataras/server-benchmarks. Run it against your own workload. ### How do I install Iris? Install Iris by running go get github.com/kataras/iris/v14@latest inside a Go module, then import github.com/kataras/iris/v14. Iris requires Go 1.27 or later and nothing else: no code generator, no CLI to install first, and no runtime beyond the binary the Go toolchain produces. The major version stays in the import path under the standard Go module rules, so a v12 and a v14 dependency can sit in one build while a codebase moves across. The API reference is published at pkg.go.dev/github.com/kataras/iris/v14, the documentation is at iris-go.com, and the repository carries 65 runnable examples grouped by topic, from getting started and routing through dependency injection, controllers, authentication, and testing. ### How do I upgrade from Iris v12 to v14? Upgrading changes the import path from github.com/kataras/iris/v12 to github.com/kataras/iris/v14, and every v12 name that v14 removed fails at compile time, so the compiler writes the task list for you. The repository ships a migration guide mapping each removed name to its replacement, and describes the work as bringing a v12 application forward in one sitting. Three defaults are stricter than they were in v12 and are worth reading before you deploy: session cookies ship with Secure and SameSite=Lax, JWT is read from the Authorization header only, and CORS will not advertise credentials alongside a wildcard origin. Packages moved as well, with sessions, i18n, view, and websocket now under middleware, and the view engine implementations in the separate iris-contrib/views repository. Go 1.27 or later is required. The guide is at https://github.com/kataras/iris/blob/main/MIGRATION.md. ## About Hellenic Identity ### What is Hellenic Identity? Hellenic Identity is a self-hosted OAuth 2.0 identity server built by Hellenic Development. It runs as one Go binary against one PostgreSQL database, on a domain you control, and it ships an admin panel for the users, their schema and the OAuth2 client. Users are described once in the server's config with typed custom attributes, groups are saved filters over those attributes, and access tokens are JWTs signed with Ed25519 that any service can verify against the server's public JWKS. It is built for organisations that cannot send identity data to a vendor's cloud. The Go SDK ships with the Iris web framework, and the C# SDK is on NuGet as Hellenic.Identity.SDK. ### Which OAuth 2.0 grants does Hellenic Identity support? Hellenic Identity supports 4 OAuth 2.0 grants: authorization code, device code with a built-in consent screen, password, and refresh token. Access tokens are JWTs signed with Ed25519 by default, and refresh tokens can be encrypted with AES-GCM on top. Lifetimes are set per client for the code, the access token and the refresh token. Bumping a client's version invalidates every token it ever issued, which is how mass revocation works: no list of revoked tokens to keep, one number to change. ### How do applications verify a Hellenic Identity token? Applications verify a Hellenic Identity token in one of two ways. The first is local: fetch the server's public keys once from /.well-known/jwks.json and check the Ed25519 signature on every request without another round trip, which is what the Go and C# SDKs do at startup. The second is to ask the server: POST the token to /oauth2/token/introspect and read back its claims, such as the username, the client id and the expiry. The server can also mint a token with extra claims for a session that needs them, which the panel and the SDKs call enrichment. ### Does Hellenic Identity support OpenID Connect? Hellenic Identity does not support OpenID Connect in this version. It is an OAuth 2.0 server with a JWKS endpoint, token introspection and token enrichment. The OpenID Connect layer exists in the source but is not switched on, so there is no userinfo endpoint and no ID token yet. An application that needs sign-in, a verified access token and the user's attributes gets those from the OAuth 2.0 endpoints and the user API today, and the same protocol code keeps working when the layer is enabled. ### What does Hellenic Identity need to run? Hellenic Identity needs one Go binary and one PostgreSQL database. The binary is built on the Iris web framework, creates its tables on first start, and prints the server token and the encryption key that the admin panel asks for. Dragonfly or Redis is optional and only used for response caching, which the panel can switch off. There is no Docker image and no hosted edition: you run it on your own server, on your own domain, and the panel at id.hellenic.dev talks to it from your browser and stores nothing anywhere else. ## About Patrol ### What is Patrol? Patrol is an open-source Go library for event production and delivery. A producer buffers the events a program sends it and flushes them in batches, when the batch fills or when the buffer interval elapses, to 7 kinds of destination: Slack, Discord, Twilio for SMS and WhatsApp, email over SMTP, a SQL table, an HTTP client that forwards to a remote patrol server, and that server itself. Each destination runs independently, so a Slack outage cannot stop the email going out, and every failure reaches one OnError callback carrying the name of the integration that produced it. Patrol needs Go 1.27 or later and nothing else, is licensed BSD-3-Clause, and is published as the Go module github.com/kataras/patrol. The current release is 0.0.7. ### Is Patrol Cloud available yet? Not yet. The Patrol library is the part that ships today: it is public, tagged 0.0.7, and it installs into a Go program in one command. Patrol Cloud, the hosted service that will receive events, store them per project and show them in a console at https://patrol.hellenic.dev, is still being built. There is no sign-up, no account and no API key to ask for. Until it opens, you get the same delivery by running the integrations inside your own process, or by running a patrol server of your own and pointing your other services at it with a patrol client. Both paths use the code Patrol Cloud will use, so a program written against the library now keeps working when the hosted service arrives. Changing the base URL and adding a key is the whole migration. ### How do I install Patrol? Install Patrol by running go get github.com/kataras/patrol@latest inside a Go module, then import github.com/kataras/patrol. Go 1.27 or later is the only requirement: no agent running beside your process, no code generator, and no CLI to install first. Build a flusher for each destination you want, hand them to NewProducer, and call SendEvent from anywhere in your program. NewException captures a filtered stacktrace with editor deep links, and Close drains whatever is still buffered before the process exits. The API reference is at pkg.go.dev/github.com/kataras/patrol, and the repository carries runnable examples for every integration with a sample configuration file beside each one. ### Is Patrol free to use? Yes. Patrol is open source under the BSD-3-Clause license, which permits commercial use, modification, and redistribution as long as the copyright notice travels with the code. There is no paid tier, no license key, and no cap on how many events a process delivers. The source is at github.com/kataras/patrol and the API reference is on pkg.go.dev. Patrol Cloud, the hosted service still in development, will publish its own prices when it opens, and the library stays free either way. Running the integrations yourself is not a trial or a reduced edition: it is the same code, and it is what Patrol Cloud runs internally. ### Can I run my own Patrol server? Yes, and that is the intended way to use Patrol before Patrol Cloud opens. NewServer returns an http.Handler holding the integrations you configured, so one service owns the Slack token and the SMTP credentials and the rest of your fleet never sees them. Every other service builds a client with NewClient, points it at that server, and sends events the same way it would send them to Slack directly. The server accepts a POST with a JSON array of events and answers a GET with OK for connectivity checks, both protected by a username and password pair or by a bearer token you list. It answers the sender before it delivers, so a slow integration never slows the caller down. ## Contact - Email: contact@hellenic.dev - Contact form: https://hellenic.dev/get-in-touch/ - LinkedIn: https://www.linkedin.com/company/hellenic-development - X: https://twitter.com/MakisMaropoulos - GitHub: https://github.com/hellenic-development