# π§ 300+ Advanced Open-Source Repositories for Professional & Senior Engineers [2026]
No fluff, no beginner detours - just the repositories senior engineers actually reach for. π§ βοΈ
Distributed systems, cloud-native infra, the full LLM-engineering stack, and a serious security arsenal.
---
> **A deep, battle-tested catalog of 300+ production-grade repositories for senior software engineers, staff engineers, SREs, platform & infrastructure engineers, security engineers, ML/AI engineers, and data engineers.**
>
> **π― Audience: Professionals & Advanced.** This is the "senior engineer" companion to the [280+ comprehensive list](../README.md) (for everyone) and the [150 essentials list](../awesome-learn-to-code/README.md) (for beginners & students). Everything here assumes you already ship software - it's about depth: distributed systems, cloud-native infra, observability, LLM engineering, and offensive/defensive security.
> **ποΈ Curated on 21 July 2026** from live research across GitHub Trending, the CNCF landscape, Papers-we-Love, and practitioner communities. Every link is a real, canonical GitHub repository.
---
## π Table of Contents
- [ποΈ System Design & Distributed Systems](#-system-design--distributed-systems)
- [ποΈ Databases & Storage Engines](#-databases--storage-engines)
- [π Data Engineering & Pipelines](#-data-engineering--pipelines)
- [π‘ Message Queues, Streaming & Event-Driven](#-message-queues-streaming--event-driven)
- [β‘ LLM Serving & Inference](#-llm-serving--inference)
- [𧩠LLM App Frameworks & Orchestration](#-llm-app-frameworks--orchestration)
- [π€ AI Agents & Multi-Agent Systems](#-ai-agents--multi-agent-systems)
- [π RAG, Vector Databases & Agent Memory](#-rag-vector-databases--agent-memory)
- [ποΈ Model Fine-Tuning & Training](#-model-fine-tuning--training)
- [π§ͺ LLM Evaluation, Observability & Guardrails](#-llm-evaluation-observability--guardrails)
- [π¦ MLOps & Experiment Tracking](#-mlops--experiment-tracking)
- [βοΈ DevOps & Infrastructure-as-Code](#-devops--infrastructure-as-code)
- [βΈοΈ Kubernetes & Cloud-Native](#-kubernetes--cloud-native)
- [π Observability & SRE](#-observability--sre)
- [𧡠Backend Frameworks & API Tooling](#-backend-frameworks--api-tooling)
- [πΈοΈ Networking, Proxies & Service Mesh](#-networking-proxies--service-mesh)
- [ποΈ Performance, Load Testing & Chaos](#-performance-load-testing--chaos)
- [ποΈ Platform Engineering & Internal Developer Platforms](#-platform-engineering--internal-developer-platforms)
- [π‘οΈ Security, AppSec & Offensive/Defensive](#-security-appsec--offensivedefensive)
- [π¦ Systems Languages, Runtimes & Containers](#-systems-languages-runtimes--containers)
- [π¨ Advanced Frontend & Build Tooling](#-advanced-frontend--build-tooling)
- [π± Advanced Mobile](#-advanced-mobile)
- [π 2026 Breakouts (Pro Picks)](#-2026-breakouts-pro-picks)
---
## ποΈ System Design & Distributed Systems
- [System Design Primer](https://github.com/donnemartin/system-design-primer) - The canonical open-source guide to scaling systems, with deep treatment of caching, sharding, and consistency trade-offs.
- [System Design 101](https://github.com/ByteByteGoHq/system-design-101) - ByteByteGo's visual explainers of real-world architectures used as shared vocabulary in design reviews.
- [System Design (karanpratapsingh)](https://github.com/karanpratapsingh/system-design) - A book-length treatment of scalability, distributed patterns, and failure modes beyond interview trivia.
- [Awesome Scalability](https://github.com/binhnguyennus/awesome-scalability) - A curated firehose of engineering blog posts and papers on reliable, performant large-scale systems.
- [Awesome System Design Resources](https://github.com/ashishps1/awesome-system-design-resources) - Concise reference of distributed-systems building blocks and case studies kept current for practitioners.
- [Distributed Systems Topics](https://github.com/Sairyss/distributed-systems-topics) - A dense index of microservices, scalability, and reliability topics with authoritative sources.
- [Jepsen](https://github.com/jepsen-io/jepsen) - Kyle Kingsbury's framework for verifying distributed database consistency under partitions; the industry standard for correctness testing.
- [PingCAP Talent Plan](https://github.com/pingcap/talent-plan) - Hands-on courses that have you build a real distributed KV store and SQL layer in Rust/Go.
- [Papers We Love](https://github.com/papers-we-love/papers-we-love) - A community repo of foundational CS papers (consensus, storage, concurrency) for first principles.
- [Awesome Distributed Systems](https://github.com/theanalyst/awesome-distributed-systems) - A tightly curated list of the seminal talks, papers, and books on distributed computing.
- [Awesome Software Architecture](https://github.com/mehdihadeli/awesome-software-architecture) - Deep, categorized resources on DDD, CQRS, event sourcing, and clean/hexagonal architecture.
- [Awesome Distributed System Projects](https://github.com/roma-glushko/awesome-distributed-system-projects) - Real open-source distributed systems organized for studying production implementations.
## ποΈ Databases & Storage Engines
- [PostgreSQL](https://github.com/postgres/postgres) - The official mirror of the world's most advanced open-source RDBMS, the default backend for serious systems.
- [CockroachDB](https://github.com/cockroachdb/cockroach) - A distributed SQL database with serializable isolation and survivability, prized for global consistency without manual sharding.
- [TiDB](https://github.com/pingcap/tidb) - A horizontally scalable HTAP database with MySQL compatibility for mixed transactional/analytical workloads.
- [Vitess](https://github.com/vitessio/vitess) - The MySQL sharding and clustering system that powers YouTube and Slack.
- [YugabyteDB](https://github.com/yugabyte/yugabyte-db) - A Postgres-compatible distributed SQL database for resilient, geo-distributed deployments.
- [ScyllaDB](https://github.com/scylladb/scylladb) - A C++ Cassandra-compatible store with a shard-per-core architecture delivering predictable low latency.
- [TiKV](https://github.com/tikv/tikv) - A CNCF-graduated distributed transactional key-value store (Raft + Percolator).
- [Apache Cassandra](https://github.com/apache/cassandra) - The battle-tested wide-column store for write-heavy, multi-datacenter workloads.
- [ClickHouse](https://github.com/ClickHouse/ClickHouse) - The fastest open-source columnar OLAP database, the default for real-time analytics over billions of rows.
- [DuckDB](https://github.com/duckdb/duckdb) - An in-process analytical database ("SQLite for analytics") ubiquitous in data pipelines.
- [Apache Doris](https://github.com/apache/doris) - A high-performance real-time MPP analytical database for sub-second queries.
- [TimescaleDB](https://github.com/timescale/timescaledb) - A Postgres extension turning it into a scalable time-series database with hypertables.
- [InfluxDB](https://github.com/influxdata/influxdb) - A purpose-built time-series database and platform for metrics and events at scale.
- [QuestDB](https://github.com/questdb/questdb) - A high-throughput time-series database with SQL and nanosecond timestamps.
- [VictoriaMetrics](https://github.com/VictoriaMetrics/VictoriaMetrics) - A fast, cost-efficient time-series database and long-term Prometheus store.
- [PgBouncer](https://github.com/pgbouncer/pgbouncer) - The lightweight Postgres connection pooler, essentially mandatory for serverless and high-concurrency backends.
- [Patroni](https://github.com/zalando/patroni) - The de facto template for Postgres high availability and automated failover.
- [RocksDB](https://github.com/facebook/rocksdb) - Facebook's embeddable LSM-tree key-value engine that serves as the storage layer for countless databases.
- [LevelDB](https://github.com/google/leveldb) - Google's classic embedded key-value store, still studied as the reference LSM implementation.
- [BadgerDB](https://github.com/dgraph-io/badger) - A fast pure-Go embeddable LSM key-value store for local persistence.
- [Pebble](https://github.com/cockroachdb/pebble) - CockroachDB's RocksDB-inspired storage engine in Go, a clean reference for modern LSM design.
- [etcd](https://github.com/etcd-io/etcd) - The distributed, strongly-consistent key-value store that backs Kubernetes.
- [Redis](https://github.com/redis/redis) - The ubiquitous in-memory data structure server for caching, queues, and real-time data.
- [Valkey](https://github.com/valkey-io/valkey) - The Linux Foundation's Redis fork that became the community standard after Redis's license change.
- [Dragonfly](https://github.com/dragonflydb/dragonfly) - A multi-threaded, Redis/Memcached-compatible in-memory store with far higher per-node throughput.
- [libSQL](https://github.com/tursodatabase/libsql) - Turso's open-contribution SQLite fork adding replication and server features.
- [Turso Database](https://github.com/tursodatabase/turso) - A ground-up SQLite rewrite in Rust with concurrent writes and async I/O.
- [SurrealDB](https://github.com/surrealdb/surrealdb) - A multi-model (document/graph/relational) database in Rust designed to collapse the backend data layer.
- [Neo4j](https://github.com/neo4j/neo4j) - The leading native graph database for connected-data workloads like fraud detection and knowledge graphs.
## π Data Engineering & Pipelines
- [Apache Airflow](https://github.com/apache/airflow) - The most widely deployed workflow orchestrator for authoring and scheduling data pipelines.
- [Dagster](https://github.com/dagster-io/dagster) - An asset-oriented orchestrator with strong typing, testing, and lineage that pros favor over Airflow.
- [Prefect](https://github.com/PrefectHQ/prefect) - Pythonic, dynamic workflow orchestration with a modern hybrid execution model.
- [dbt Core](https://github.com/dbt-labs/dbt-core) - The transformation-layer standard for version-controlled, tested SQL analytics engineering.
- [Apache Spark](https://github.com/apache/spark) - The dominant distributed engine for large-scale batch and streaming data processing.
- [Apache Flink](https://github.com/apache/flink) - The reference stateful stream-processing engine for low-latency, exactly-once pipelines.
- [Polars](https://github.com/pola-rs/polars) - A blazing-fast, multi-threaded Rust DataFrame library increasingly replacing pandas for performance.
- [Apache Arrow](https://github.com/apache/arrow) - The columnar in-memory format and IPC standard underpinning the modern data stack.
- [dlt](https://github.com/dlt-hub/dlt) - A Python-first, code-driven EL library for loading data into warehouses with schema evolution.
- [Apache Iceberg](https://github.com/apache/iceberg) - An open table format bringing ACID, schema evolution, and time travel to data lakes.
- [Delta Lake](https://github.com/delta-io/delta) - A storage layer adding ACID transactions and versioning to lakehouse architectures.
- [Apache Hudi](https://github.com/apache/hudi) - A lakehouse platform bringing incremental processing and upserts to data lakes.
- [Great Expectations](https://github.com/great-expectations/great_expectations) - A framework for validating, documenting, and profiling data quality.
- [SQLFluff](https://github.com/sqlfluff/sqlfluff) - The dialect-flexible SQL linter and formatter for keeping analytics code clean in CI.
- [Airbyte](https://github.com/airbytehq/airbyte) - An open-source ELT platform with the largest catalog of data connectors.
- [Ibis](https://github.com/ibis-project/ibis) - A portable Python dataframe API that compiles to 20+ SQL and dataframe backends.
- [Apache Beam](https://github.com/apache/beam) - A unified programming model for portable batch and streaming pipelines across runners.
- [PRQL](https://github.com/PRQL/prql) - A pipelined relational query language that compiles to SQL for more maintainable analytics.
## π‘ Message Queues, Streaming & Event-Driven
- [Apache Kafka](https://github.com/apache/kafka) - The dominant distributed event-streaming platform and the backbone of most event-driven architectures.
- [Redpanda](https://github.com/redpanda-data/redpanda) - A Kafka-API-compatible streaming engine in C++ with no ZooKeeper/JVM, valued for latency and simplicity.
- [NATS Server](https://github.com/nats-io/nats-server) - A tiny, ultra-fast messaging system with JetStream persistence, popular for microservice and edge messaging.
- [RabbitMQ](https://github.com/rabbitmq/rabbitmq-server) - The versatile, protocol-rich message broker that remains a default for reliable task queues.
- [Apache Pulsar](https://github.com/apache/pulsar) - A streaming/messaging platform with separated compute and storage, multi-tenancy, and geo-replication.
- [Temporal](https://github.com/temporalio/temporal) - A durable execution platform for writing fault-tolerant, long-running workflows as ordinary code.
- [Cadence](https://github.com/cadence-workflow/cadence) - Uber's durable workflow orchestration engine (Temporal's predecessor) for long-running business logic.
- [Apache RocketMQ](https://github.com/apache/rocketmq) - A high-throughput, low-latency distributed messaging and streaming platform proven at Alibaba scale.
- [Debezium](https://github.com/debezium/debezium) - The standard change-data-capture platform for streaming database changes into Kafka and beyond.
- [Redpanda Connect (Benthos)](https://github.com/redpanda-data/connect) - A declarative stream processor for glue-code-free data plumbing and transformation.
- [RisingWave](https://github.com/risingwavelabs/risingwave) - A Postgres-compatible streaming database for building real-time materialized views in SQL.
- [NSQ](https://github.com/nsqio/nsq) - A realtime distributed messaging platform in Go designed for simplicity, scale, and no SPOF.
## β‘ LLM Serving & Inference
- [vLLM](https://github.com/vllm-project/vllm) - The high-throughput PagedAttention serving engine that is the de facto open-source production inference standard.
- [SGLang](https://github.com/sgl-project/sglang) - A fast serving framework with RadixAttention prefix caching, prized for structured generation and low latency.
- [llama.cpp](https://github.com/ggml-org/llama.cpp) - The C/C++ GGUF inference engine enabling efficient CPU/GPU/Apple-silicon local LLM deployment.
- [Ollama](https://github.com/ollama/ollama) - The most popular local model runner, giving a clean CLI/API over llama.cpp with model management.
- [TensorRT-LLM](https://github.com/NVIDIA/TensorRT-LLM) - NVIDIA's compiler and runtime for squeezing maximum throughput out of Tensor Core GPUs.
- [Text Generation Inference](https://github.com/huggingface/text-generation-inference) - Hugging Face's production Rust/Python inference server that powered many hosted endpoints.
- [LMDeploy](https://github.com/InternLM/lmdeploy) - InternLM's toolkit for compressing and serving LLMs with high-performance TurboMind kernels.
- [LightLLM](https://github.com/ModelTC/lightllm) - A pure-Python, token-attention serving framework favored for lightweight, hackable high-concurrency inference.
- [KTransformers](https://github.com/kvcache-ai/ktransformers) - A framework for running huge MoE models on limited GPU/CPU hardware via offloading and kernel injection.
- [Aphrodite Engine](https://github.com/PygmalionAI/aphrodite-engine) - A vLLM-derived engine focused on high-throughput serving with broad quantization support.
- [LocalAI](https://github.com/mudler/LocalAI) - A drop-in OpenAI-compatible API for self-hosted text, image, and audio models across many backends.
- [Text Embeddings Inference](https://github.com/huggingface/text-embeddings-inference) - A blazing-fast Rust server for embedding and reranker models, a key RAG-stack component.
- [NVIDIA Dynamo](https://github.com/ai-dynamo/dynamo) - A datacenter-scale distributed inference framework with disaggregated prefill/decode serving.
- [llamafile](https://github.com/Mozilla-Ocho/llamafile) - A Mozilla project that packages a model and runtime into a single cross-platform executable.
- [exo](https://github.com/exo-explore/exo) - Runs a single large model as a distributed cluster across everyday devices.
## 𧩠LLM App Frameworks & Orchestration
- [LangChain](https://github.com/langchain-ai/langchain) - The most widely adopted LLM app framework and integration layer for chains, tools, and retrieval.
- [LlamaIndex](https://github.com/run-llama/llama_index) - The data framework specializing in ingestion, indexing, and retrieval for production RAG.
- [LangGraph](https://github.com/langchain-ai/langgraph) - A graph-based runtime for stateful, human-in-the-loop agent workflows; the enterprise orchestration favorite.
- [DSPy](https://github.com/stanfordnlp/dspy) - A Stanford framework that programs and auto-optimizes prompts/weights instead of hand-tuning strings.
- [Haystack](https://github.com/deepset-ai/haystack) - deepset's modular, production-grade pipeline framework for RAG and search applications.
- [LiteLLM](https://github.com/BerriAI/litellm) - A universal proxy/SDK giving one OpenAI-format interface to 100+ model providers with cost tracking.
- [Dify](https://github.com/langgenius/dify) - An open-source, visual low-code platform for building and operating LLM apps and agents.
- [Langflow](https://github.com/langflow-ai/langflow) - A visual builder for RAG and multi-agent flows with a Python backend.
- [Flowise](https://github.com/FlowiseAI/Flowise) - A drag-and-drop UI for assembling LLM orchestration flows.
- [Semantic Kernel](https://github.com/microsoft/semantic-kernel) - Microsoft's enterprise SDK for planners, plugins, and agent orchestration across .NET/Python.
- [Pydantic AI](https://github.com/pydantic/pydantic-ai) - A type-safe, Pydantic-native agent framework for structured, validated LLM outputs.
- [Mastra](https://github.com/mastra-ai/mastra) - A TypeScript-first framework for agents, workflows, and RAG, popular in the JS/edge ecosystem.
- [Portkey Gateway](https://github.com/Portkey-AI/gateway) - A fast open-source AI gateway with routing, fallbacks, caching, and observability.
- [Guidance](https://github.com/guidance-ai/guidance) - A constrained-generation language for interleaving control flow with token-level structure.
- [Outlines](https://github.com/dottxt-ai/outlines) - A structured generation library guaranteeing JSON-schema/regex/CFG-valid LLM outputs.
- [BAML](https://github.com/BoundaryML/baml) - A domain-specific language for building reliable, type-safe, testable LLM functions.
- [Vercel AI SDK](https://github.com/vercel/ai) - The TypeScript toolkit for building AI apps and agents across many providers.
## π€ AI Agents & Multi-Agent Systems
- [AutoGen](https://github.com/microsoft/autogen) - Microsoft's multi-agent conversation framework and a research staple.
- [CrewAI](https://github.com/crewAIInc/crewAI) - A lean framework for orchestrating role-playing agent "crews" with minimal boilerplate.
- [OpenHands](https://github.com/All-Hands-AI/OpenHands) - An autonomous software-engineering agent platform (formerly OpenDevin) that writes, runs, and tests code.
- [MetaGPT](https://github.com/geekan/MetaGPT) - A multi-agent framework that simulates a software company with role-based SOPs.
- [OpenAI Agents SDK](https://github.com/openai/openai-agents-python) - A lightweight, provider-agnostic framework for multi-agent workflows with handoffs and guardrails.
- [smolagents](https://github.com/huggingface/smolagents) - Hugging Face's minimalist code-writing agent library.
- [Google ADK](https://github.com/google/adk-python) - Google's Agent Development Kit for building, evaluating, and deploying production agents.
- [CAMEL](https://github.com/camel-ai/camel) - A research framework for communicative and role-playing multi-agent systems at scale.
- [Agno](https://github.com/agno-agi/agno) - A high-performance framework/runtime for agents with memory, knowledge, and tools (formerly Phidata).
- [SWE-agent](https://github.com/SWE-agent/SWE-agent) - Princeton's benchmark-topping agent that autonomously resolves GitHub issues.
- [Skyvern](https://github.com/Skyvern-AI/skyvern) - Automates browser workflows using vision-LLMs that read pages via screenshots.
- [Stagehand](https://github.com/browserbase/stagehand) - An AI browser-automation SDK blending natural language with code for reliable web agents.
- [LiveKit Agents](https://github.com/livekit/agents) - A framework for building real-time, low-latency voice and multimodal AI agents.
- [ChatDev](https://github.com/OpenBMB/ChatDev) - A multi-agent virtual software company for collaborative code generation.
- [E2B](https://github.com/e2b-dev/E2B) - Open-source secure cloud sandboxes for running AI-generated code.
## π RAG, Vector Databases & Agent Memory
- [Milvus](https://github.com/milvus-io/milvus) - A cloud-native vector database engineered for billion-scale similarity search.
- [Qdrant](https://github.com/qdrant/qdrant) - A Rust vector database praised for fast filtered search and simple production operations.
- [Weaviate](https://github.com/weaviate/weaviate) - A vector database with strong hybrid search and a built-in module ecosystem.
- [Chroma](https://github.com/chroma-core/chroma) - A developer-friendly embedding database that is the default for fast RAG prototyping.
- [pgvector](https://github.com/pgvector/pgvector) - The Postgres extension that adds vector search, the pragmatic default for teams already on Postgres.
- [LanceDB](https://github.com/lancedb/lancedb) - An embedded, multimodal vector database built on the columnar Lance format.
- [FAISS](https://github.com/facebookresearch/faiss) - Meta's foundational library for efficient similarity search and vector clustering.
- [Marqo](https://github.com/marqo-ai/marqo) - An end-to-end vector search engine with first-class multimodal embedding generation.
- [RAGFlow](https://github.com/infiniflow/ragflow) - An end-to-end RAG engine with deep document understanding and agentic context construction.
- [GraphRAG](https://github.com/microsoft/graphrag) - Microsoft's knowledge-graph-based RAG approach for global reasoning over large corpora.
- [Graphiti](https://github.com/getzep/graphiti) - Builds real-time, temporally-aware knowledge graphs for agent memory.
- [cognee](https://github.com/topoteretes/cognee) - A memory/knowledge-graph layer that replaces ad-hoc RAG with structured pipelines.
- [txtai](https://github.com/neuml/txtai) - An all-in-one embeddings database for semantic search, RAG, and LLM workflows.
- [Unstructured](https://github.com/Unstructured-IO/unstructured) - A preprocessing library that turns PDFs, HTML, and docs into clean, chunked LLM data.
- [Mem0](https://github.com/mem0ai/mem0) - A universal memory layer giving agents persistent user/session/agent-scoped memory.
- [Letta](https://github.com/letta-ai/letta) - An agent framework (formerly MemGPT) treating memory like an OS with self-managed context.
## ποΈ Model Fine-Tuning & Training
- [Transformers](https://github.com/huggingface/transformers) - The foundational model library and hub interface underpinning most modern LLM work.
- [Unsloth](https://github.com/unslothai/unsloth) - Custom Triton kernels that make LoRA/QLoRA fine-tuning 2x+ faster with far less VRAM.
- [Axolotl](https://github.com/axolotl-ai-cloud/axolotl) - A config-driven fine-tuning framework favored for reproducible, production training pipelines.
- [LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory) - A unified, GUI-friendly fine-tuning platform supporting 100+ models and every major PEFT method.
- [PEFT](https://github.com/huggingface/peft) - Hugging Face's parameter-efficient fine-tuning library (LoRA, adapters, prompt tuning) used everywhere.
- [TRL](https://github.com/huggingface/trl) - A post-training toolkit unifying SFT, DPO, GRPO, and RLHF trainers on the HF stack.
- [Megatron-LM](https://github.com/NVIDIA/Megatron-LM) - NVIDIA's reference for large-scale tensor/pipeline-parallel transformer pretraining.
- [DeepSpeed](https://github.com/microsoft/DeepSpeed) - Microsoft's ZeRO-based training/inference optimization library for training massive models.
- [Accelerate](https://github.com/huggingface/accelerate) - A thin abstraction for launching PyTorch training across multi-GPU/TPU/mixed-precision setups.
- [Diffusers](https://github.com/huggingface/diffusers) - The standard toolkit for training and running diffusion models for image/audio/video.
- [OpenRLHF](https://github.com/OpenRLHF/OpenRLHF) - A Ray-based, high-performance RLHF framework for scalable PPO/GRPO alignment training.
- [verl](https://github.com/verl-project/verl) - Volcano Engine's flexible RL post-training library (HybridFlow), a 2026 favorite.
- [ms-swift](https://github.com/modelscope/ms-swift) - ModelScope's framework for fine-tuning and deploying 400+ LLMs/MLLMs.
- [XTuner](https://github.com/InternLM/xtuner) - An efficient, flexible toolkit for fine-tuning LLMs and multimodal models.
- [LitGPT](https://github.com/Lightning-AI/litgpt) - 20+ from-scratch LLM implementations with pretrain/finetune/deploy recipes.
## π§ͺ LLM Evaluation, Observability & Guardrails
- [Langfuse](https://github.com/langfuse/langfuse) - An open-source LLM engineering platform for tracing, evals, and prompt management; the observability leader.
- [Ragas](https://github.com/explodinggradients/ragas) - A reference-free RAG evaluation library that pioneered faithfulness/relevancy metrics.
- [DeepEval](https://github.com/confident-ai/deepeval) - "Pytest for LLMs" with 50+ metrics and CI/CD-friendly unit testing of model outputs.
- [Phoenix](https://github.com/Arize-ai/phoenix) - Arize's OpenTelemetry-native, self-hostable tracing and evaluation tool for LLM/agent apps.
- [Guardrails](https://github.com/guardrails-ai/guardrails) - Adds input/output validators and structured-output guarantees around LLM calls.
- [NeMo Guardrails](https://github.com/NVIDIA/NeMo-Guardrails) - NVIDIA's toolkit for adding programmable safety/topical rails to conversational apps.
- [OpenAI Evals](https://github.com/openai/evals) - The widely-used framework and registry for building and running model evaluations.
- [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) - EleutherAI's de facto harness for academic LLM benchmarking.
- [promptfoo](https://github.com/promptfoo/promptfoo) - A CLI/library for test-driven prompt evaluation and LLM red-teaming/security scanning.
- [lighteval](https://github.com/huggingface/lighteval) - Hugging Face's lightweight, all-in-one evaluation suite across multiple backends.
- [Giskard](https://github.com/Giskard-AI/giskard) - A testing framework that scans ML/LLM systems for vulnerabilities, bias, and robustness issues.
- [LLM Guard](https://github.com/protectai/llm-guard) - A security toolkit for defending against prompt injection and data leakage.
- [OpenLLMetry](https://github.com/traceloop/openllmetry) - OpenTelemetry-based instrumentation for end-to-end LLM app observability.
- [Helicone](https://github.com/Helicone/helicone) - An open-source LLM observability proxy for logging, caching, and cost monitoring.
- [Opik](https://github.com/comet-ml/opik) - Comet's open-source platform for tracing, evaluating, and monitoring LLM applications.
## π¦ MLOps & Experiment Tracking
- [MLflow](https://github.com/mlflow/mlflow) - The de facto open standard for experiment tracking, model registry, and GenAI evaluation.
- [Ray](https://github.com/ray-project/ray) - A distributed compute framework (Train/Tune/Serve/Data) underpinning much large-scale ML infra.
- [BentoML](https://github.com/bentoml/BentoML) - A framework for packaging models and building/serving production inference APIs.
- [KServe](https://github.com/kserve/kserve) - A Kubernetes-native, standardized model inference platform with autoscaling and vLLM support.
- [ZenML](https://github.com/zenml-io/zenml) - An extensible MLOps/LLMOps framework for portable, reproducible pipelines across stacks.
- [Weights & Biases](https://github.com/wandb/wandb) - The industry-standard experiment tracking, artifacts, and model management SDK.
- [Aim](https://github.com/aimhubio/aim) - An open-source, self-hostable experiment tracker for high-volume run comparison.
- [DVC](https://github.com/iterative/dvc) - Git-based data and model version control for reproducible ML pipelines.
- [Kubeflow](https://github.com/kubeflow/kubeflow) - The Kubernetes-native ML platform for pipelines, training, and serving at scale.
- [Flyte](https://github.com/flyteorg/flyte) - A strongly-typed, Kubernetes-native workflow orchestrator for scalable ML/data pipelines.
- [Metaflow](https://github.com/Netflix/metaflow) - Netflix's human-centric framework for building and managing real-world data science projects.
- [Feast](https://github.com/feast-dev/feast) - The leading open-source feature store for serving consistent features to training and inference.
- [SkyPilot](https://github.com/skypilot-org/skypilot) - Runs training/serving/batch jobs across any cloud or Kubernetes with cost and spot optimization.
- [OpenLLM](https://github.com/bentoml/OpenLLM) - Run and deploy open LLMs as OpenAI-compatible endpoints with one command.
## βοΈ DevOps & Infrastructure-as-Code
- [Terraform](https://github.com/hashicorp/terraform) - The dominant multi-cloud IaC tool; fluency here is table stakes for infra engineers.
- [OpenTofu](https://github.com/opentofu/opentofu) - The Linux Foundation's open-source Terraform fork, increasingly adopted after the BSL relicense.
- [Pulumi](https://github.com/pulumi/pulumi) - IaC using real languages (TypeScript, Go, Python), valued where teams want tests, loops, and abstractions.
- [Ansible](https://github.com/ansible/ansible) - The agentless configuration-management and automation engine that remains a fleet-management workhorse.
- [Crossplane](https://github.com/crossplane/crossplane) - A CNCF-graduated control plane that manages cloud infra as Kubernetes resources.
- [Packer](https://github.com/hashicorp/packer) - HashiCorp's tool for building identical machine images across clouds.
- [Vault](https://github.com/hashicorp/vault) - The industry-standard secrets management and dynamic-credentials system for zero-trust infra.
- [Consul](https://github.com/hashicorp/consul) - Service discovery, service mesh, and distributed KV for connecting and securing services.
- [Nomad](https://github.com/hashicorp/nomad) - A simple, flexible workload orchestrator scheduling containers and non-container workloads.
- [SOPS](https://github.com/getsops/sops) - The standard for encrypting secrets in Git (YAML/JSON) with KMS/age.
- [Terragrunt](https://github.com/gruntwork-io/terragrunt) - A thin Terraform wrapper for keeping large multi-environment configs DRY.
- [Infracost](https://github.com/infracost/infracost) - Shows cloud cost diffs in pull requests so teams catch expensive infra changes before merge.
- [LocalStack](https://github.com/localstack/localstack) - A local AWS cloud emulator for developing and testing cloud apps without cost or latency.
- [Dagger](https://github.com/dagger/dagger) - A programmable, containerized CI/CD engine that lets you write pipelines as code and run them anywhere.
- [Kubespray](https://github.com/kubernetes-sigs/kubespray) - Production-grade Ansible playbooks for deploying self-managed Kubernetes clusters.
- [Salt](https://github.com/saltstack/salt) - An event-driven automation and configuration-management platform for large infrastructures.
- [AWX](https://github.com/ansible/awx) - The upstream open-source project behind Ansible Automation Platform.
- [Earthly](https://github.com/earthly/earthly) - Reproducible, container-native builds that run identically locally and in CI.
## βΈοΈ Kubernetes & Cloud-Native
- [Kubernetes](https://github.com/kubernetes/kubernetes) - The container orchestration platform that defines cloud-native infrastructure.
- [Helm](https://github.com/helm/helm) - The de facto Kubernetes package manager for templating and releasing complex applications.
- [Kustomize](https://github.com/kubernetes-sigs/kustomize) - Template-free, overlay-based Kubernetes configuration management built into kubectl.
- [Argo CD](https://github.com/argoproj/argo-cd) - The leading declarative GitOps continuous-delivery controller for Kubernetes.
- [Flux](https://github.com/fluxcd/flux2) - A CNCF-graduated GitOps toolkit for keeping clusters continuously reconciled with Git.
- [Argo Workflows](https://github.com/argoproj/argo-workflows) - A Kubernetes-native workflow engine for orchestrating parallel jobs and data/ML pipelines.
- [Argo Rollouts](https://github.com/argoproj/argo-rollouts) - Progressive delivery (canary, blue-green) with automated analysis and rollback.
- [Istio](https://github.com/istio/istio) - The most feature-complete service mesh, providing mTLS, traffic management, and telemetry.
- [Linkerd](https://github.com/linkerd/linkerd2) - A lightweight, security-first service mesh valued for operational simplicity.
- [k3s](https://github.com/k3s-io/k3s) - A certified lightweight Kubernetes distribution ideal for edge, IoT, and CI.
- [kind](https://github.com/kubernetes-sigs/kind) - Runs Kubernetes clusters in Docker, the standard for local testing and CI of cluster software.
- [Rancher](https://github.com/rancher/rancher) - A complete platform for managing fleets of Kubernetes clusters across on-prem and cloud.
- [Cluster API](https://github.com/kubernetes-sigs/cluster-api) - Declarative, Kubernetes-style APIs for managing the lifecycle of clusters themselves.
- [KubeVirt](https://github.com/kubevirt/kubevirt) - Runs traditional VMs alongside containers on Kubernetes.
- [Open Policy Agent](https://github.com/open-policy-agent/opa) - A general-purpose policy engine (Rego) for authorization and admission control.
- [Kyverno](https://github.com/kyverno/kyverno) - A Kubernetes-native policy engine using YAML instead of a new language.
- [Harbor](https://github.com/goharbor/harbor) - A CNCF-graduated container registry with vulnerability scanning, signing, and replication.
- [Tekton Pipelines](https://github.com/tektoncd/pipeline) - A Kubernetes-native, reusable framework for building cloud-native CI/CD pipelines.
- [cert-manager](https://github.com/cert-manager/cert-manager) - Automates issuance and renewal of TLS certificates in Kubernetes.
- [Kubebuilder](https://github.com/kubernetes-sigs/kubebuilder) - The SDK for building Kubernetes APIs and operators with CRDs.
- [Knative Serving](https://github.com/knative/serving) - Serverless scale-to-zero and request-driven autoscaling primitives on Kubernetes.
- [KEDA](https://github.com/kedacore/keda) - Event-driven autoscaling on queue depth, Kafka lag, and dozens of other sources.
- [Velero](https://github.com/vmware-tanzu/velero) - Backup, restore, and disaster recovery for Kubernetes resources and persistent volumes.
- [Longhorn](https://github.com/longhorn/longhorn) - A CNCF distributed block storage system for Kubernetes with replication and snapshots.
- [Rook](https://github.com/rook/rook) - A storage operator that turns Ceph into self-managing storage on Kubernetes.
## π Observability & SRE
- [Prometheus](https://github.com/prometheus/prometheus) - The de facto cloud-native metrics and alerting system; its data model defines modern monitoring.
- [Grafana](https://github.com/grafana/grafana) - The universal observability dashboard for metrics, logs, and traces across any backend.
- [Loki](https://github.com/grafana/loki) - A cost-efficient, label-based log aggregation system that indexes metadata instead of full text.
- [Tempo](https://github.com/grafana/tempo) - A high-scale distributed tracing backend that needs only object storage to run.
- [Mimir](https://github.com/grafana/mimir) - Horizontally scalable, long-term Prometheus storage handling a billion+ active series.
- [Pyroscope](https://github.com/grafana/pyroscope) - A continuous-profiling database to debug CPU/memory down to the line of code in production.
- [OpenTelemetry Collector](https://github.com/open-telemetry/opentelemetry-collector) - The vendor-neutral pipeline for receiving, processing, and exporting all telemetry.
- [Jaeger](https://github.com/jaegertracing/jaeger) - The CNCF-graduated distributed tracing system for root-causing latency across microservices.
- [Thanos](https://github.com/thanos-io/thanos) - Adds global query, unlimited retention, and HA to Prometheus via object storage.
- [Vector](https://github.com/vectordotdev/vector) - A blazing-fast Rust observability data pipeline for collecting, transforming, and routing telemetry.
- [Fluent Bit](https://github.com/fluent/fluent-bit) - A lightweight, high-throughput log/metric processor and forwarder for edge and containers.
- [Netdata](https://github.com/netdata/netdata) - Real-time, per-second infrastructure monitoring with zero-config auto-discovery.
- [SigNoz](https://github.com/SigNoz/signoz) - An OpenTelemetry-native, all-in-one APM and observability platform (open-source Datadog).
- [OpenObserve](https://github.com/openobserve/openobserve) - A petabyte-scale observability platform with dramatically lower storage cost.
- [Parca](https://github.com/parca-dev/parca) - Always-on eBPF-based continuous profiling to cut CPU and memory usage across a fleet.
- [Pixie](https://github.com/pixie-io/pixie) - eBPF-powered, no-instrumentation observability for Kubernetes with automatic protocol tracing.
- [Coroot](https://github.com/coroot/coroot) - An eBPF-based observability tool that auto-generates service maps and SLO-driven insights.
- [Grafana OnCall](https://github.com/grafana/oncall) - Open-source on-call scheduling and incident escalation, an alternative to PagerDuty.
- [Keep](https://github.com/keephq/keep) - An open-source AIOps and alert-correlation platform to tame alert fatigue.
- [Robusta](https://github.com/robusta-dev/robusta) - Automation and enrichment that adds context and remediation to Kubernetes alerts.
## 𧡠Backend Frameworks & API Tooling
- [Gin](https://github.com/gin-gonic/gin) - The most popular Go HTTP framework, valued for speed and a minimal API.
- [Echo](https://github.com/labstack/echo) - A high-performance, minimalist Go web framework with strong middleware ergonomics.
- [Fiber](https://github.com/gofiber/fiber) - An Express-inspired Go framework built on fasthttp for maximum throughput.
- [chi](https://github.com/go-chi/chi) - A lightweight, idiomatic Go router that composes cleanly with the standard library.
- [gRPC-Go](https://github.com/grpc/grpc-go) - The Go implementation of gRPC, the default for high-performance internal APIs.
- [Connect-Go](https://github.com/connectrpc/connect-go) - A simpler, browser-compatible RPC framework interoperable with gRPC.
- [sqlc](https://github.com/sqlc-dev/sqlc) - Generates fully type-safe Go from raw SQL, a favorite for teams avoiding heavyweight ORMs.
- [Ent](https://github.com/ent/ent) - Facebook's graph-based, schema-as-code ORM for Go with strong static typing.
- [GORM](https://github.com/go-gorm/gorm) - The full-featured Go ORM that remains the most-used data access layer.
- [golang-migrate](https://github.com/golang-migrate/migrate) - Database schema migrations as a library and CLI for every major database.
- [Tokio](https://github.com/tokio-rs/tokio) - The asynchronous runtime that anchors virtually all production Rust network services.
- [Axum](https://github.com/tokio-rs/axum) - The Tokio team's ergonomic, modular web framework, now the mainstream Rust choice.
- [Actix Web](https://github.com/actix/actix-web) - An extremely fast, mature Rust web framework topping throughput benchmarks.
- [SQLx](https://github.com/launchbadge/sqlx) - Async, compile-time-checked SQL for Rust with no DSL.
- [Serde](https://github.com/serde-rs/serde) - The foundational Rust serialization framework used by essentially every Rust backend.
- [Tower](https://github.com/tower-rs/tower) - Composable middleware powering timeouts, retries, and load-shedding in Rust services.
- [Spring Boot](https://github.com/spring-projects/spring-boot) - The dominant JVM framework for production microservices and enterprise backends.
- [Quarkus](https://github.com/quarkusio/quarkus) - A Kubernetes-native Java stack with fast startup and low memory via AOT compilation.
- [Micronaut](https://github.com/micronaut-projects/micronaut-core) - A modern JVM framework using compile-time DI to avoid reflection.
- [Ktor](https://github.com/ktorio/ktor) - JetBrains' asynchronous Kotlin framework for connected server and client applications.
- [Netty](https://github.com/netty/netty) - The asynchronous event-driven network framework underpinning much JVM networking.
- [NestJS](https://github.com/nestjs/nest) - A structured, TypeScript-first Node framework bringing DI and modular architecture to backends.
- [Fastify](https://github.com/fastify/fastify) - A high-throughput, low-overhead Node framework with schema-based validation.
- [Hono](https://github.com/honojs/hono) - An ultrafast, runtime-agnostic web framework that runs on Workers, Deno, Bun, and Node.
- [tRPC](https://github.com/trpc/trpc) - End-to-end type-safe APIs without codegen, popular in TypeScript monorepos.
- [Prisma](https://github.com/prisma/prisma) - A type-safe ORM and toolkit that has become a default data layer for TypeScript backends.
- [gRPC](https://github.com/grpc/grpc) - The core multi-language RPC framework for efficient, contract-first service communication.
- [Protocol Buffers](https://github.com/protocolbuffers/protobuf) - Google's language-neutral serialization format, the schema backbone of gRPC.
- [Buf](https://github.com/bufbuild/buf) - Modern Protobuf tooling with linting, breaking-change detection, and a schema registry.
- [OpenAPI Specification](https://github.com/OAI/OpenAPI-Specification) - The standard for describing REST APIs that drives codegen, docs, and contract testing.
## πΈοΈ Networking, Proxies & Service Mesh
- [Envoy](https://github.com/envoyproxy/envoy) - The CNCF-graduated L7 proxy at the heart of most service meshes and modern edge infra.
- [Traefik](https://github.com/traefik/traefik) - A cloud-native reverse proxy and ingress with automatic service discovery and Let's Encrypt.
- [Caddy](https://github.com/caddyserver/caddy) - A Go web server/proxy with automatic HTTPS by default, valued for operational simplicity.
- [Cilium](https://github.com/cilium/cilium) - eBPF-based networking, security, and observability; the leading modern CNI.
- [Kong](https://github.com/Kong/kong) - A high-performance, plugin-extensible API gateway built on Nginx/OpenResty.
- [HAProxy](https://github.com/haproxy/haproxy) - The battle-hardened, ultra-reliable TCP/HTTP load balancer for demanding traffic.
- [CoreDNS](https://github.com/coredns/coredns) - The CNCF-graduated, plugin-based DNS server providing service discovery inside Kubernetes.
- [MetalLB](https://github.com/metallb/metallb) - A load-balancer implementation for bare-metal Kubernetes clusters.
- [Pingora](https://github.com/cloudflare/pingora) - Cloudflare's Rust framework for fast, programmable proxies, serving 40M+ requests/second.
- [Tetragon](https://github.com/cilium/tetragon) - eBPF-based runtime security observability and enforcement with low overhead.
- [Hubble](https://github.com/cilium/hubble) - Cilium's eBPF-powered network, service, and security observability layer.
- [Kubeshark](https://github.com/kubeshark/kubeshark) - An API traffic analyzer ("Wireshark for Kubernetes") for real-time network visibility.
## ποΈ Performance, Load Testing & Chaos
- [k6](https://github.com/grafana/k6) - A developer-centric, scriptable load-testing tool (JavaScript) that has become the modern standard.
- [Locust](https://github.com/locustio/locust) - Distributed load testing with scenarios defined in Python.
- [Vegeta](https://github.com/tsenart/vegeta) - A versatile HTTP load-testing tool for constant-rate attacks with rich reporting.
- [Gatling](https://github.com/gatling/gatling) - A high-performance load-testing framework with an expressive DSL for the JVM ecosystem.
- [FlameGraph](https://github.com/brendangregg/FlameGraph) - Brendan Gregg's flame graph visualization, the canonical way to interpret CPU profiles.
- [bpftrace](https://github.com/bpftrace/bpftrace) - A high-level tracing language for Linux eBPF, indispensable for deep production performance analysis.
- [BCC](https://github.com/iovisor/bcc) - A toolkit of eBPF-based tools for advanced kernel tracing, profiling, and networking observability.
- [async-profiler](https://github.com/async-profiler/async-profiler) - A low-overhead sampling profiler for the JVM without safepoint bias.
- [wrk2](https://github.com/giltene/wrk2) - A constant-throughput wrk fork with corrected latency measurement.
- [hyperfine](https://github.com/sharkdp/hyperfine) - A command-line benchmarking tool with statistical rigor for comparing program performance.
- [Chaos Mesh](https://github.com/chaos-mesh/chaos-mesh) - A CNCF cloud-native chaos-engineering platform for injecting faults into Kubernetes.
- [Litmus](https://github.com/litmuschaos/litmus) - A CNCF chaos-engineering framework for practicing resilience testing.
- [Toxiproxy](https://github.com/Shopify/toxiproxy) - Shopify's TCP proxy for deterministically simulating network failures and latency in tests.
## ποΈ Platform Engineering & Internal Developer Platforms
- [Backstage](https://github.com/backstage/backstage) - Spotify's CNCF developer portal and software catalog, the de facto standard for IDPs.
- [Kratix](https://github.com/syntasso/kratix) - A platform orchestrator that delivers self-service APIs between portals and infrastructure.
- [Score](https://github.com/score-spec/spec) - A platform-agnostic workload specification decoupling developer intent from environment config.
- [Porter](https://github.com/porter-dev/porter) - An internal developer platform that deploys apps to your own cloud with a Heroku-like experience.
- [Devtron](https://github.com/devtron-labs/devtron) - An open-source Kubernetes-native software delivery platform unifying CI/CD, GitOps, and security.
- [werf](https://github.com/werf/werf) - A CNCF GitOps CLI tying Git, build, and Helm deploy into reproducible delivery pipelines.
- [Kusion](https://github.com/KusionStack/kusion) - An intent-driven platform orchestrator for building app-centric internal developer platforms.
## π‘οΈ Security, AppSec & Offensive/Defensive
- [Semgrep](https://github.com/semgrep/semgrep) - Fast, rule-based static analysis (SAST) with pattern-as-code, widely adopted in CI security gates.
- [TruffleHog](https://github.com/trufflesecurity/trufflehog) - Scans repos, history, and systems for leaked credentials and verifies them live.
- [Gitleaks](https://github.com/gitleaks/gitleaks) - A fast, CI-friendly secrets scanner that is a standard pre-commit and pipeline check.
- [Trivy](https://github.com/aquasecurity/trivy) - The all-in-one scanner for container images, IaC, SBOMs, and secrets, ubiquitous in DevSecOps.
- [Grype](https://github.com/anchore/grype) - A fast vulnerability scanner for container images and filesystems, often paired with Syft.
- [Syft](https://github.com/anchore/syft) - Generates SBOMs (SPDX/CycloneDX) from images and filesystems, foundational to supply-chain security.
- [Cosign](https://github.com/sigstore/cosign) - Sigstore's tool for signing and verifying container images and artifacts.
- [Checkov](https://github.com/bridgecrewio/checkov) - Policy-as-code static analysis for Terraform, CloudFormation, and Kubernetes.
- [OWASP ZAP](https://github.com/zaproxy/zaproxy) - The leading open-source DAST web app scanner and proxy for runtime vulnerabilities.
- [Nuclei](https://github.com/projectdiscovery/nuclei) - Template-driven, high-speed vulnerability scanning, the standard for automated detection.
- [Subfinder](https://github.com/projectdiscovery/subfinder) - A fast passive subdomain enumeration tool, a staple of recon workflows.
- [httpx](https://github.com/projectdiscovery/httpx) - A fast, multi-purpose HTTP toolkit for probing and fingerprinting hosts at scale.
- [naabu](https://github.com/projectdiscovery/naabu) - A fast, reliable port scanner in Go built for large-scale recon pipelines.
- [Amass](https://github.com/owasp-amass/amass) - OWASP's in-depth attack-surface mapping and external asset discovery framework.
- [sqlmap](https://github.com/sqlmapproject/sqlmap) - The definitive automated SQL injection detection and exploitation tool.
- [Sliver](https://github.com/BishopFox/sliver) - Bishop Fox's cross-platform adversary-emulation C2 framework, a modern Cobalt Strike alternative.
- [Havoc](https://github.com/HavocFramework/Havoc) - A modern, malleable post-exploitation C2 framework popular in red-team engagements.
- [BloodHound](https://github.com/SpecterOps/BloodHound) - Graph-based analysis of Active Directory/Azure attack paths.
- [NetExec](https://github.com/Pennyw0rth/NetExec) - The maintained successor to CrackMapExec for automating network and AD assessments.
- [PayloadsAllTheThings](https://github.com/swisskyrepo/PayloadsAllTheThings) - The go-to reference of payloads and bypass techniques for web app pentesting.
- [SecLists](https://github.com/danielmiessler/SecLists) - The essential collection of wordlists for fuzzing, brute-forcing, and discovery.
- [hashcat](https://github.com/hashcat/hashcat) - The world's fastest GPU-accelerated password recovery tool.
- [Metasploit Framework](https://github.com/rapid7/metasploit-framework) - The canonical exploitation framework for penetration testing and exploit development.
- [PEASS-ng](https://github.com/carlospolop/PEASS-ng) - The definitive privilege-escalation enumeration scripts (linPEAS/winPEAS).
- [Prowler](https://github.com/prowler-cloud/prowler) - Multi-cloud security assessments (AWS/Azure/GCP/K8s) mapped to CIS and compliance benchmarks.
- [kube-bench](https://github.com/aquasecurity/kube-bench) - Checks Kubernetes clusters against the CIS Kubernetes Benchmark.
- [Kubescape](https://github.com/kubescape/kubescape) - A Kubernetes security platform for misconfiguration, RBAC, and image scanning.
- [Cloud Custodian](https://github.com/cloud-custodian/cloud-custodian) - A YAML rules engine for cloud governance, cost, and security guardrails at scale.
- [ScoutSuite](https://github.com/nccgroup/ScoutSuite) - NCC Group's multi-cloud security auditing tool.
- [Pacu](https://github.com/RhinoSecurityLabs/pacu) - An open-source AWS exploitation framework for offensive cloud security testing.
- [Wazuh](https://github.com/wazuh/wazuh) - An open-source SIEM/XDR platform for threat detection, integrity monitoring, and compliance.
- [Velociraptor](https://github.com/Velocidex/velociraptor) - Endpoint visibility and digital forensics/incident response at fleet scale.
- [TheHive](https://github.com/TheHive-Project/TheHive) - A scalable security incident response platform for SOC collaboration.
- [Falco](https://github.com/falcosecurity/falco) - The CNCF runtime security engine detecting anomalous behavior via eBPF/syscalls.
- [CrowdSec](https://github.com/crowdsecurity/crowdsec) - A collaborative, behavior-based intrusion prevention system with crowd-sourced threat intel.
- [OSSF Scorecard](https://github.com/ossf/scorecard) - Automatically scores open-source projects' security posture for supply-chain risk.
## π¦ Systems Languages, Runtimes & Containers
- [Go](https://github.com/golang/go) - The reference implementation of the Go language and standard library, essential for backend engineers.
- [Rust](https://github.com/rust-lang/rust) - The Rust compiler and standard library, the foundation of the memory-safe systems ecosystem.
- [Zig](https://github.com/ziglang/zig) - A simple, explicit systems language and toolchain increasingly used as the build layer under other projects.
- [Awesome Go](https://github.com/avelino/awesome-go) - The definitive curated index of Go libraries and tools.
- [Awesome Rust](https://github.com/rust-unofficial/awesome-rust) - The canonical curated catalog of Rust libraries, applications, and systems tooling.
- [Go Project Layout](https://github.com/golang-standards/project-layout) - The widely referenced conventions for structuring real-world Go applications.
- [Uber Go Style Guide](https://github.com/uber-go/guide) - Uber's battle-tested Go style guide adopted by many teams as their standard.
- [fasthttp](https://github.com/valyala/fasthttp) - A zero-allocation HTTP implementation for Go with order-of-magnitude throughput gains.
- [conc](https://github.com/sourcegraph/conc) - Sourcegraph's structured-concurrency toolkit for correct, ergonomic goroutine management.
- [Rayon](https://github.com/rayon-rs/rayon) - Data-parallelism for Rust that makes CPU-bound code parallel with a one-line change.
- [clap](https://github.com/clap-rs/clap) - The full-featured, ergonomic command-line argument parser powering most Rust CLIs.
- [Firecracker](https://github.com/firecracker-microvm/firecracker) - AWS's Rust microVM tech powering Lambda/Fargate, the reference for secure lightweight virtualization.
- [containerd](https://github.com/containerd/containerd) - The CNCF-graduated industry-standard container runtime underpinning Docker and Kubernetes.
- [runc](https://github.com/opencontainers/runc) - The OCI reference low-level runtime for spawning containers.
- [Podman](https://github.com/containers/podman) - A daemonless, rootless container engine that has become the secure Docker alternative on Linux.
- [CRI-O](https://github.com/cri-o/cri-o) - A lightweight, Kubernetes-dedicated container runtime implementing the CRI.
- [Wasmtime](https://github.com/bytecodealliance/wasmtime) - The Bytecode Alliance's fast, secure WebAssembly/WASI runtime and Component Model reference.
- [Wasmer](https://github.com/wasmerio/wasmer) - A leading universal WebAssembly runtime for running Wasm across languages and platforms.
- [Bun](https://github.com/oven-sh/bun) - An all-in-one JS runtime, bundler, and package manager built for speed.
- [Deno](https://github.com/denoland/deno) - A secure-by-default, TypeScript-first runtime with a modern standard library.
- [Ruff](https://github.com/astral-sh/ruff) - Astral's extremely fast Rust-written Python linter and formatter.
- [uv](https://github.com/astral-sh/uv) - Astral's ultra-fast Rust package/project manager rapidly replacing pip/Poetry.
- [Biome](https://github.com/biomejs/biome) - A Rust-based, all-in-one formatter and linter positioned as a fast Prettier/ESLint replacement.
- [SWC](https://github.com/swc-project/swc) - A Rust platform for fast TS/JS compilation and transformation used inside many toolchains.
- [typescript-go](https://github.com/microsoft/typescript-go) - Microsoft's native Go port of the TypeScript compiler (the basis of TS 7) targeting ~10x speedups.
## π¨ Advanced Frontend & Build Tooling
- [React](https://github.com/facebook/react) - The core UI library whose Server Components and compiler define much of modern frontend.
- [Next.js](https://github.com/vercel/next.js) - The dominant React meta-framework for SSR/SSG/RSC production apps.
- [Svelte](https://github.com/sveltejs/svelte) - A compiler-based framework (v5 runes) delivering minimal-runtime, high-performance UIs.
- [SvelteKit](https://github.com/sveltejs/kit) - The official application framework for building Svelte apps.
- [SolidJS](https://github.com/solidjs/solid) - A fine-grained reactive library offering React-like ergonomics with near-vanilla performance.
- [Astro](https://github.com/withastro/astro) - A content-focused framework with islands architecture that ships zero JS by default.
- [Nuxt](https://github.com/nuxt/nuxt) - The intuitive Vue meta-framework for hybrid-rendered, full-stack applications.
- [Qwik](https://github.com/QwikDev/qwik) - A resumable framework eliminating hydration for instant-loading web apps.
- [Vite](https://github.com/vitejs/vite) - The near-universal dev server and build tool that became the frontend tooling backbone.
- [Turborepo](https://github.com/vercel/turborepo) - A Rust-based high-performance build system with caching for JS/TS monorepos.
- [Nx](https://github.com/nrwl/nx) - An extensible monorepo build system with computation caching and rich generators.
- [Rolldown](https://github.com/rolldown/rolldown) - A Rust-based, Rollup-compatible bundler becoming Vite's unified production bundler.
- [Oxc](https://github.com/oxc-project/oxc) - A Rust suite of ultra-fast JS/TS tools (parser, linter, resolver, minifier).
- [Rspack](https://github.com/web-infra-dev/rspack) - A Rust bundler with a webpack-compatible API for drop-in speed on large codebases.
- [TanStack Query](https://github.com/TanStack/query) - The standard async-state/data-fetching library for React and other frameworks.
- [TanStack Router](https://github.com/TanStack/router) - A fully type-safe router with first-class search-param and data-loading APIs.
- [React Router](https://github.com/remix-run/react-router) - The routing standard that absorbed Remix into a full-stack React framework.
- [shadcn/ui](https://github.com/shadcn-ui/ui) - The copy-in, unstyled-by-default component collection that redefined design systems.
- [Tailwind CSS](https://github.com/tailwindlabs/tailwindcss) - The utility-first CSS framework whose v4 Oxide engine is a modern styling default.
## π± Advanced Mobile
- [React Native](https://github.com/facebook/react-native) - The cross-platform native app framework whose New Architecture closed the performance gap.
- [Expo](https://github.com/expo/expo) - The recommended production toolchain for React Native with builds, OTA updates, and native modules.
- [Compose Multiplatform](https://github.com/JetBrains/compose-multiplatform) - JetBrains' declarative UI sharing Compose across Android, iOS, desktop, and web.
- [The Composable Architecture](https://github.com/pointfreeco/swift-composable-architecture) - Point-Free's state-management library for testable, scalable SwiftUI apps.
- [React Native Reanimated](https://github.com/software-mansion/react-native-reanimated) - The high-performance native-thread animation library essential to serious RN apps.
- [Now in Android](https://github.com/android/nowinandroid) - Google's reference app showcasing modern Android architecture and Compose best practices.
- [Capacitor](https://github.com/ionic-team/capacitor) - A native runtime for building cross-platform apps with web tech and full native API access.
- [Tamagui](https://github.com/tamagui/tamagui) - A universal UI kit and optimizing compiler for React Native + web from one codebase.
## π 2026 Breakouts (Pro Picks)
_Repositories that broke out or went viral between late 2025 and mid-2026 - the newest additions to a senior engineer's radar._
- [OpenClaw](https://github.com/openclaw/openclaw) - Self-hosted personal AI assistant that wires any LLM into WhatsApp/Telegram/Slack/Signal; went from ~9k to 250k+ stars in early 2026, the fastest-growing repo of the year.
- [Codex CLI](https://github.com/openai/codex) - OpenAI's Rust-rewritten open-source terminal coding agent (~95k+ stars in 2025).
- [Gemini CLI](https://github.com/google-gemini/gemini-cli) - Google's open-source terminal AI agent embedding Gemini, one of the most-watched dev tools of 2025-2026.
- [OpenCode](https://github.com/sst/opencode) - A fast, provider-flexible terminal coding agent that became a dominant open CLI agent in 2026.
- [Crush](https://github.com/charmbracelet/crush) - Charm's polished Go TUI coding agent with multi-model and MCP support.
- [Cline](https://github.com/cline/cline) - The leading autonomous VS Code coding agent with human-in-the-loop approval and MCP tools.
- [Kilo Code](https://github.com/Kilo-Org/kilocode) - An all-in-one agentic VS Code extension merging Roo/Cline ideas, a fast-growing 2025-2026 fork.
- [Goose](https://github.com/block/goose) - Block's extensible on-machine agentic coding assistant, now under Linux Foundation governance.
- [Pi (pi-mono)](https://github.com/badlogic/pi-mono) - A radically minimal terminal coding agent + harness toolkit that blew up on a sub-1k-token system prompt.
- [MCP Servers](https://github.com/modelcontextprotocol/servers) - The reference collection of Model Context Protocol servers, the standard for connecting agents to tools/data.
- [GitHub MCP Server](https://github.com/github/github-mcp-server) - GitHub's official MCP server for repo/issue/PR/Actions access from agents.
- [Playwright MCP](https://github.com/microsoft/playwright-mcp) - Microsoft's browser-automation MCP server, the go-to way to give agents a real browser.
- [gpt-oss](https://github.com/openai/gpt-oss) - OpenAI's first open-weight models (120b/20b, Apache-2.0) since GPT-2, a shock release in Aug 2025.
- [DeepSeek-V3.2-Exp](https://github.com/deepseek-ai/DeepSeek-V3.2-Exp) - A MoE model introducing DeepSeek Sparse Attention for cheap long-context inference.
- [Kimi-K2](https://github.com/MoonshotAI/Kimi-K2) - Moonshot's trillion-parameter MoE built for agentic, long-horizon tasks.
- [Qwen3](https://github.com/QwenLM/Qwen3) - Alibaba's Apache-2.0 dense + MoE family with hybrid thinking modes, a default open-weight backbone.
- [GLM-4.5](https://github.com/zai-org/GLM-4.5) - Z.ai's agentic/reasoning/coding foundation models, a top open coding/agent model.
- [Crawl4AI](https://github.com/unclecode/crawl4ai) - An LLM-friendly web crawler that hit #1 GitHub trending as RAG data-ingestion demand exploded.
- [Firecrawl](https://github.com/mendableai/firecrawl) - Turns entire websites into LLM-ready Markdown/structured data via a scrape-and-crawl API.
- [Docling](https://github.com/docling-project/docling) - Parses PDFs and office docs into structured formats for GenAI (LF AI & Data).
- [Motia](https://github.com/MotiaDev/motia) - A unified backend framework merging APIs, jobs, queues, workflows, and AI agents.
- [Better Auth](https://github.com/better-auth/better-auth) - A comprehensive, framework-agnostic TypeScript authentication library.
- [Ghostty](https://github.com/ghostty-org/ghostty) - A fast, GPU-accelerated, native cross-platform terminal emulator.
- [Zed](https://github.com/zed-industries/zed) - A high-performance multiplayer code editor from the creators of Atom.
---
## π€ Contributing
Found a production-grade repo that belongs here? See [CONTRIBUTING.md](../CONTRIBUTING.md) and open a pull request. Advanced, real-world, actively-maintained repositories only.
## π§ The Three Lists
| List | Audience | Link |
| ---------------------- | -------------------------------- | ----------------------------------------------------- |
| **150 Essentials** | Beginners & Students | [awesome-learn-to-code](../awesome-learn-to-code/README.md) |
| **280+ Comprehensive** | Everyone | [Main README](../README.md) |
| **300+ Advanced** | Professionals & Senior Engineers | _(you are here)_ |
Built for people who ship. βοΈ
If a repo here saved you a war-room at 3 a.m., pay it forward with a PR.
π Licensed under Apache 2.0 Β· π§ Curated by md8-habibullah
If this helped, star β and share it with your team.