Open source observability and logging, installed and run

Observability and logging tools you can run yourself. The list is ordered by whether each project came up when Argusic installed it fresh, with the recording of that attempt one click away.

Tested between and . Each row shows its own test date; a project can change after that day.

86 of 93 tested projects run. 6 more waiting for a test.

In short: 76 of the 93 tested projects started as-is on a fresh machine: jaeger, witr, vector, VictoriaMetrics, zipkin, thanos, openllmetry, and hertzbeat, and 68 more. 10 more started once a stand-in replaced a service they expect, such as a database: open-multi-agent, kube-state-metrics, odigos, MyPerf4J, subtrace, future-agi, GoModel, and kindling, and 2 more. 7 could not be verified: calico, ecapture, beyla, pup, trinity, Tracekit, and coze-loop; the log shows where each one stopped.

Measured by Argusic on a fresh machine every time. Every number links to its evidence.

#projectverdictArgusic Scorelanguagestarstested on
1jaegerRuns100 / 100Go23,441

CNCF Jaeger, a Distributed Tracing Platform

What the test found: jaeger (v2) builds from source, passes 5244 unit tests with 0 lint issues, and the all-in-one binary starts and serves HTTP on port 16686. 17 minutes.

2witrRuns100 / 100Go23,062

Why is this running? Trace any process, port, container, or file back to what started it - CLI + TUI.

What the test found: witr builds, passes all 8 test packages (including -race), and traces real processes in the container. 4 minutes.

3vectorRuns100 / 100Rust22,936

A high-performance observability data pipeline.

What the test found: A preserved Vector binary reports version 0.59.0-dev-custom-3ce1054, validates a YAML config, and successfully processed three sentinel log records through demo_logs, remap, and console with no stderr output. 78 minutes.

4VictoriaMetricsRuns100 / 100Go18,170

VictoriaMetrics: fast, cost-effective monitoring solution and time series database

What the test found: VictoriaMetrics single binary builds, runs as an HTTP server on port 8428 returning valid responses, and the full unit test suite (131 packages) plus integration tests pass cleanly. 17 minutes.

5zipkinRuns100 / 100Java17,475

Zipkin is a distributed tracing system

What the test found: Zipkin server 3.6.2-SNAPSHOT builds and runs on JDK 21, serving HTTP on port 9411, accepting spans via API v2 with InMemoryStorage, and passing all 817 unit tests. 10 minutes.

6thanosRuns100 / 100Go14,286

Highly available Prometheus setup with long term storage capabilities. A CNCF Incubating project.

What the test found: The thanos binary builds and runs: its query component answered HTTP 200 on /-/, served thanos_build_info metrics, resolved a real gRPC store.LabelNames call, and the filesystem object store round-tripped uploaded data, while the unit-test suite passed in every package except pkg/reloader, which cannot run in this... 21 minutes.

7openllmetryRuns100 / 100Python7,547

Open-source observability for your GenAI or LLM application, based on OpenTelemetry

What the test found: Traceloop SDK and all its instrumentation packages install, import, build, and pass their full test suites without any errors. 2 minutes.

8hertzbeatRuns100 / 100Java7,472

An AI-powered next-generation open source real-time observability system.

What the test found: HertzBeat server builds, its full Maven test suite passes except one Docker-only integration test, and the assembled server runs with working JWT login and authenticated API on port 1157 in this container. 37 minutes.

9btraceRuns100 / 100Java5,994

Production-safe dynamic tracing and diagnostics for Java applications, attach to live JVMs with no restart or recompilation.

What the test found: BTrace 3.0.0-SNAPSHOT builds from source, passes all 1552 unit and integration tests, produces distribution archives (tar.gz, zip, deb, rpm, sdkman), and the btrace CLI prints v3.0.0-SNAPSHOT (dev) with full help output. 34 minutes.

10cost-xrayRuns100 / 100Python5,797

See what Claude Code and Codex actually send to the API, and what each part costs.

What the test found: All 223 tests pass, ruff lint passes, the mitmproxy addon loads, captures, redacts secrets, deduplicates request blocks, and materializes cost-attribution output against a mock Anthropic API endpoint. The TUI modules (XrayApp, CostXrayTUI) import successfully. 9 minutes.

11kenerRuns100 / 100TypeScript5,257

Stunning status pages, batteries included!

What the test found: Kener v4.1.3 is built, serves a status page on port 3000 with SQLite and Redis, all 136 tests pass, health endpoint confirms database and Redis connectivity. 14 minutes.

12ayaRuns100 / 100Rust4,925

Aya is an eBPF library for the Rust programming language, built with a focus on developer experience and operability.

What the test found: The aya Rust workspace builds cleanly and its full host-side test suite passes (507 tests, 0 failures) after fixing a /proc/pid/maps parser bug where 64-bit overlayfs inode numbers overflowed a u32 field; CLI tools aya-tool and xtask answer their help commands, while eBPF program compilation and kernel integration... 11 minutes.

13CPA-Manager-PlusRuns100 / 100Go4,100

A self-hosted CPA / CLIProxyAPI management panel and AI gateway observability dashboard for requests, usage, cost, quota, failures, and account health.

What the test found: CPA Manager Plus web app builds to a single-file index.html, Go manager server compiles and listens on :18317 with /health returning 200, all 4156 frontend and all Go backend tests pass without errors. 24 minutes.

14mtailRuns100 / 100Go4,033

extract internal monitoring data from application logs for collection in a timeseries database

What the test found: mtail built from source with Go 1.23.2, all 23 test packages pass, one-shot log parsing correctly counts lines from a test log file, and the binary responds to --help with version 4c7e0e1. 3 minutes.

15kialiRuns100 / 100Go3,649

Kiali project, observability for the Istio service mesh

What the test found: Kiali backend binary built at /home/runner/go/bin/kiali (176 MB), frontend UI built at frontend/build/, kiali --help displays usage, 83/83 Go unit tests pass. 37 minutes.

16express-status-monitorRuns100 / 100JavaScript3,610

🚀 Realtime Monitoring solution for Node.js/Express.js apps, inspired by status.github.com, sponsored by https://dynobase.dev

What the test found: All dependencies installed, all 24 unit tests pass, and the example Express server starts and serves the live monitoring dashboard at / with HTTP 200. 3 minutes.

17quillRuns100 / 100C++3,093

Ultra-low-latency asynchronous C++17 logging and metrics library for performance-critical applications

What the test found: The Quill C++17 logging library builds with cmake and g++13, all 179 test executables pass, and the simple_setup example runs producing a formatted log line. 11 minutes.

18TokenTrackerRuns100 / 100JavaScript2,261

Local-first AI token usage & cost tracker for 31 coding tools incl. Claude Code, Codex, Cursor, Gemini & DeepSeek Harness, with native apps. Never reads...

What the test found: All 3052 CLI tests pass with 0 failures; the tokentracker serve command starts and returns HTTP 200 on port 7680. 26 minutes.

19evlogRuns100 / 100TypeScript1,948

Digging through logs is not observability. It's hope, wide events, structured errors, TypeScript-first, every runtime.

What the test found: All 35 workspace packages install and build on Node 24. Test suite passes (1905/1907 tests, 90/91 files), lint and typecheck clean. CLI binary outputs help text. Package dist exposes 35 named exports via ESM. 6 minutes.

20trpc-agent-goRuns100 / 100Go1,940

A Go framework for building production agent systems with graph workflows, tools, memory, A2A, AG-UI, MCP, evaluation, and observability.

What the test found: The tRPC-Agent-Go framework compiles and all 235 test suites pass under Go 1.21, the E2E tests pass under Go 1.24.4, and examples compile successfully under Go 1.24.4. 25 minutes.

21openchoreoRuns100 / 100Go1,925

OpenChoreo is an internal developer platform for Kubernetes

What the test found: OpenChoreo v1.3.0 Go project builds 4 binaries, all 229 non-e2e Go unit test packages pass, and the occ CLI reports version 1.3.0 correctly. 69 minutes.

22kubetailRuns100 / 100Go1,811

Real-time logging dashboard for Kubernetes. View logs in a terminal or a browser. Run anywhere - desktop, cluster, docker.

What the test found: Frontend test suite (596/596 passed, 33/33 files), Go backend test suite (all packages passed, 0 failures), and dashboard-ui production build all succeed; the kubetail CLI binary at bin/kubetail (86 MB) responds to --help and --version correctly. 17 minutes.

23odd-platformRuns100 / 100Java1,442

First open-source data discovery and observability platform. We make a life for data practitioners easy so you can focus on your business.

What the test found: ODD Platform UI installs, builds, and passes all 384 unit tests across 57 test files. The generated TypeScript API client compiles correctly against the OpenAPI spec. 8 minutes.

24updoRuns100 / 100Go1,148

Uptime monitoring CLI tool with alerting and advanced settings

What the test found: Updo builds with Go 1.27.4, all 10 test packages pass, the binary monitors real URLs with correct HTTP status/response time/SSL tracking displayed in simple, TUI, and JSON log modes, and the config file with 7 concurrent targets with diverse settings (assertions, headers, POST, redirect handling) executes correctly. 3 minutes.

25OrcaReplayRuns100 / 100TypeScript1,127

OrcaReplay, Time travel for AI agents. Record, replay, fork, and debug any agent run with any model. Built by the OrcaRouter.ai team.

What the test found: npm install, tsc build, all 3024 tests pass, orca CLI responds to --help, orca quickstart creates a project and replays a recorded agent run offline with no network calls. 22 minutes.

26fastapi-observabilityRuns100 / 100Python1,123

Observe FastAPI app with three pillars of observability: Traces (Tempo), Metrics (Prometheus), Logs (Loki) on Grafana through OpenTelemetry.

What the test found: FastAPI application starts with uvicorn on port 8000, serves all 7 endpoints (/ , /items/{item_id}, /io_task, /cpu_task, /random_status, /random_sleep, /error_test) and the /metrics Prometheus endpoint with correct HTTP status codes, structured JSON response bodies, and OpenTelemetry trace/span context logged to... 5 minutes.

27pintRuns100 / 100Go1,062

Prometheus rule linter/validator

What the test found: pint Prometheus rule linter builds from source with Go 1.27.4 on Ubuntu 24.04, binary responds to --help and version, all 14 test packages pass. 10 minutes.

28loongsuite-goRuns100 / 100Go907

OpenTelemetry Compile-Time Instrumentation for Golang

What the test found: The otel binary builds and runs; all non-Docker integration tests pass; the pkg module unit tests pass; the tool prints its version and usage correctly. 36 minutes.

29httpdbgRuns100 / 100Python906

A tool for Python developers to easily debug the HTTP(S) client and server requests in a Python program.

What the test found: httpdbg 2.2.1 installed and all 189 non-UI tests pass; the pyhttpdbg CLI reports version 2.2.1 and displays help. 3 minutes.

30Tracely-aiRuns100 / 100Python848

Trace-native CI/CD for AI agents: grade production traces, cluster failures, freeze them into replayable regression tests, and block PRs that regress...

What the test found: Python project installs with uv sync; 1330/1330 backend+SDK tests pass; ruff lint clean; CLI binaries (gate/replay/simulate/export) respond to --help; FastAPI app instantiates; SDK modules import correctly. 11 minutes.

31yokaiRuns100 / 100Go839

Simple, modular, and observable Go framework for backend applications.

What the test found: All 29 Go modules (config, generate, grpcserver, healthcheck, httpclient, httpserver, log, orm, sql, trace, worker, fxconfig, fxcore, fxclock, fxcron, fxgenerate, fxgrpcserver, fxhealthcheck, fxhttpclient, fxhttpserver, fxlog, fxmcpserver, fxmetrics, fxorm, fxsql, fxtrace, fxvalidator, fxworker) build successfully and... 7 minutes.

32nestjs-otelRuns100 / 100TypeScript802

OpenTelemetry (Tracing + Metrics) module for Nest framework (node.js) 🔭

What the test found: The nestjs-otel library builds (tsc -> lib/) and all 118 tests pass (82 unit + 36 e2e) on Node 22.23.5. 2 minutes.

33agentacctRuns100 / 100Python758

See what your coding agents did and what it cost. Breaks each task down into work steps, tools used, files changed, tests run, time and tokens spent...

What the test found: agentacct 0.12.5 installs via venv, CLI (agentacct --version prints 0.12.5), onboard runs successfully writing MCP config and hooks, the local API starts on 127.0.0.1:8765 and returns 200 on /health with real session usage data ingested, and the test suite passes 4887 of 4895 tests. 76 minutes.

34event-reduceRuns100 / 100TypeScript756

An algorithm to optimize database queries that run multiple times https://pubkey.github.io/event-reduce/

What the test found: The event-reduce-js JavaScript package builds and passes all 51 unit tests on Node 18, exporting calculateActionName, runAction, getSortFieldsOfQuery, and 40+ other functions. 1 minute.

35apertureRuns100 / 100Go732

Rate limiting, caching, and request prioritization for modern workloads

What the test found: Aperture controller, agent, and CLI binaries build; controller serves API on :8080, agent on :18081, agent connects to controller via etcd, flow-control SDK demo app sends requests through agent producing real metrics in Prometheus. 35 minutes.

36otel-cliRuns100 / 100Go726

OpenTelemetry command-line tool for sending events from shell scripts & similar environments

What the test found: otel-cli builds from source with go build and passes all 4 test packages (58 fixtures across 23 suites) using real in-process gRPC and HTTP OTLP servers; the binary also prints help, version, and status correctly. 2 minutes.

37opik-openclawRuns100 / 100TypeScript723

🦞 Official plugin for OpenClaw that exports agent traces to Opik. See and monitor agent behaviour, cost, tokens, errors and more.

What the test found: The @opik/opik-openclaw v0.2.17 package installs, builds, typechecks, lints, and passes all 102 unit/smoke tests under Node.js v22.14.0 via vitest. The only skipped test is the E2E test that requires a live Opik server. 5 minutes.

38ApplicationInsights-JSRuns100 / 100TypeScript687

Microsoft Application Insights SDK for JavaScript

What the test found: Rush install, full build (all 31 packages), and all 18 test suites (thousands of tests in headless Chrome) passed with zero failures. 58 minutes.

39booklogrRuns100 / 100JavaScript641

A simple, self-hosted service to keep track of your personal library 📚

What the test found: BookLogr API (Python/Flask) installed, database migrated, and running at localhost:5150 with registration, login, and OpenLibrary book search all verified working. 11 minutes.

40datacatRuns100 / 100TypeScript548

Self-hosted observability: metrics, logs, traces, dashboards and alerts. Stores data in PostgreSQL or ClickHouse. Accepts OpenTelemetry data.

What the test found: DataCat builds from source with pnpm, its 59 unit tests all pass, and the server launches on port 3000 returning HTTP 200 on /, /health, and /setup. 5 minutes.

41rootprintRuns100 / 100Svelte445

Open-source, self-hosted logs and traces with full-text search directly on S3-compatible object storage.

What the test found: Install, build, type-check, lint, format check, and all 93 API test cases pass against real PostgreSQL 16 and Quickwit 0.9.1 instances running in the container. 15 minutes.

42nerdlogRuns97.5 / 100Go1,592

Nerdlog: fast, remote-first, multi-host TUI log viewer with timeline histogram and no central server

What the test found: Nerdlog builds with go build via CGO_ENABLED=0; all unit tests pass (86 subtests across blhistory, shellescape, core, cmd/nerdlog) except the tmux-dependent E2E tests; GNU Awk was extracted from a .deb and placed in ~/.local/bin for the nerdlog_agent.sh script; clipboard is disabled due to missing X11 headers. 16 minutes.

43logfireRuns97.3 / 100Python4,566

AI observability platform for production LLM and agent systems.

What the test found: The logfire SDK version 5.0.0 is installed, its CLI reports version info, and 47 of 48 test modules pass with ~1392 tests passing, 2 skipped, and 0 failures. 45 minutes.

44retinaRuns96.7 / 100Go3,189

eBPF distributed networking observability tool for Kubernetes

What the test found: Retina project builds and all unit tests pass after installing Go 1.27.3. Two tests fail due to missing system binaries (bpftool, etcd) that cannot be installed without root. Both container binaries (retina-agent, captureworkload) compile and run. 32 minutes.

45sensu-goRuns96.7 / 100Go1,114

Simple. Scalable. Multi-cloud monitoring.

What the test found: Sensu Go builds from source, all unit tests pass across 117+ packages, and all three CLI binaries (sensu-backend, sensu-agent, sensuctl) produce valid version output. 7 minutes.

46prefectRuns96 / 100Python24,210

Prefect is a workflow orchestration framework for building resilient data pipelines in Python.

What the test found: Prefect 3.6.24 is installed in a .venv, imports cleanly, the server starts and responds on /api/health with 200, and 840 core tests pass when run with uv run pytest. 22 minutes.

47opensreRuns96 / 100Python14,304

The open-source agent for CI/CD reliability.

What the test found: OpenSRE v0.1 is installed from source and builds successfully; the CLI starts and responds to all commands (help, health, doctor, version); all 274 tests and all quality gates (lint, format, typecheck, import checks) pass. 5 minutes.

48voltagentRuns96 / 100TypeScript11,001

AI Agent Engineering Platform built on an Open Source TypeScript AI Agent Framework

What the test found: Monorepo builds (29/29 packages), lints (0 errors, 5 warnings), and all 25 packages with runnable unit tests pass (~2776 tests). The only test failure is @voltagent/e2e which requires an unreachable PostgreSQL database. Two test files with a pre-existing ESM dependency conflict are excluded. 26 minutes.

49agent-beaconRuns96 / 100Go3,585

The cross-harness, self-improving memory layer for AI agents. Join our community: https://discord.com/invite/zdNChS2fBu

What the test found: The Beacon CLI builds, runs, and responds to commands; all Go tests pass (131/131), all TypeScript/Javascript tests pass (307/307), and the built binary correctly reports its version 7e47847. 10 minutes.

50pezzoRuns96 / 100TypeScript3,287

🕹️ Open-source, developer-first LLMOps platform designed to streamline prompt design, version management, instant delivery, collaboration, troubleshooting...

What the test found: Pezzo server (v0.8.0) is running on port 3000 with GraphQL and Swagger endpoints, all 22 Prisma migrations are applied to PostgreSQL 16.4, ClickHouse 24.8.14.39 with knex migrations is operational, Redis 7.2.3 answers PONG, and the console front-end builds and serves on port 4200. 20 minutes.

51slothRuns96 / 100Go2,574

🦥 Easy and simple Prometheus SLO (service level objectives) generator

What the test found: Sloth builds from source, all unit tests pass (including lib-as-CLI integration tests), CLI generate/validate/version commands work, the web server starts with a fake backend and responds on health/metrics/UI endpoints, and the K8s controller runs in fake mode. 5 minutes.

52karporRuns96 / 100Go1,724

Intelligence for Kubernetes. World's most promising Kubernetes Visualization Tool for Developer and Platform Engineering teams.

What the test found: Karpor server binary builds, connects to etcd, listens on :7443 with TLS, returns 200 on /livez and /readyz, exposes all REST API endpoints including cluster, search, insight, resource-group-rule management, and 18 test packages pass. 12 minutes.

53qtapRuns96 / 100C1,465

Qtap: An eBPF agent that captures pre-encrypted network traffic, providing rich context about egress connections and their originating processes.

What the test found: Go 1.27.1 was downloaded and installed; go build produces a working qtap binary at /work/repo/bin/qtap that responds to --help and --version; the test suite passes on 49 of 51 packages (only pkg/egress IPv6 tests fail because this container has IPv6 disabled). 14 minutes.

54opentelemetry-go-instrumentationRuns96 / 100C1,039

OpenTelemetry Auto Instrumentation using eBPF

What the test found: OpenTelemetry Go Automatic Instrumentation builds from source, the CLI binary (80MB) compiles and prints usage text, and all 27+ test packages across cli, sdk, internal/pkg pass with 0 failures; the tool is ready for use targeting Go processes via eBPF. 7 minutes.

55skywalking-javaRuns96 / 100Java928

The Java agent for Apache SkyWalking

What the test found: Full build and test suite pass. The Java agent package (skywalking-agent.jar plus plugin/activation/bootstrap jars and config) compiles, produces a valid Premain-Class manifest, attaches correctly to a JVM via -javaagent, and all 944 unit tests pass with zero failures or errors. 45 minutes.

56golfRuns96 / 100Python842

Production-Ready MCP Server Framework • Build, deploy & scale secure AI agent infrastructure • Includes Auth, Observability, Debugger, Telemetry & Runtime •...

What the test found: golf-mcp v0.4.1 installed from source; CLI commands (init, build dev, build prod, run) work; server starts on port 3000 with streamable-http transport and responds to authenticated MCP requests; 305/306 tests pass. 4 minutes.

57opentelemetry-collectorRuns95 / 100Go7,794

OpenTelemetry Collector

What the test found: The OpenTelemetry Collector builds, validates configs, boots up, serves OTLP gRPC (4317) and HTTP (4318) endpoints, serves zpages debugging (55679), and all 98 Go module test suites pass except 2 IPv6 tests that fail due to the container having no IPv6 loopback. 10 minutes.

58superlogRuns93.5 / 100TypeScript1,490

Open-source observability tool that uses AI agents to self-heal your software

What the test found: 17 workspace packages pass typecheck; all test suites execute against real Postgres and ClickHouse servers stood up in-container -- API reports 569/674 pass in parallel (102 from concurrent-DB isolation, all pass sequentially). 27 minutes.

59grafana-operatorRuns92.7 / 100Go1,395

An operator for Grafana that installs and manages Grafana instances, Dashboards and Datasources through Kubernetes/OpenShift CRs

What the test found: Go 1.27.4 installed; 12 test suites (93 cases) pass; bin/manager builds and prints help; linter clean. 5 minutes.

60corootRuns90 / 100Go8,029

Coroot is an open-source observability and APM tool with AI-powered Root Cause Analysis. It combines metrics, logs, traces, continuous profiling, and SLO-based...

What the test found: Coroot binary builds, frontend static assets compile, all Go tests pass (18 packages ok, 0 failures), and the server starts on port 8090 responding HTTP 200 on its root and API endpoints. 12 minutes.

61agent-starter-packRuns90 / 100-6,572

Ship AI Agents to Google Cloud in minutes, not months. Production-ready templates with built-in CI/CD, evaluation, and observability.

What the test found: The agent-starter-pack v0.41.3 CLI installs, builds, and passes all 498 unit tests with the agent-starter-pack list, --help, and --version commands working correctly; uv is available at /home/runner/.local/bin/uv. 8 minutes.

62k8spacketRuns75 / 100Go1,095

k8spacket - collects TCP traffic and TLS connection metadata in the Kubernetes cluster using eBPF and visualizes in Grafana

What the test found: The k8spacket Go project builds successfully with CGO_ENABLED=0 into a static binary; all 31 test packages pass (26 testable packages across cmd, internal/broker, internal/ebpf/*, internal/modules/*, internal/thirdparty/*); the binary starts and serves HTTP /metrics on port 6676 but eBPF program loading is blocked due... 6 minutes.

63codeburnRuns66.7 / 100TypeScript11,884

Free, local tool to track AI coding token usage and cost across 37 tools and agents (Claude Code, Cursor, Codex, Gemini and more), by model, project, and task...

What the test found: CodeBurn v0.9.23 builds, passes 3546 of 3567 tests, TypeScript compiles clean, and the CLI prints help and version on demand. 8 minutes.

64hyperdxRuns64.4 / 100TypeScript10,064

Resolve production issues, fast. An open source observability platform unifying session replays, logs, metrics, traces and errors powered by ClickHouse and...

What the test found: All unit test suites pass across 4 packages with Node 22.23.1 (7120 tests, 0 failures). Lint passes with warnings only. 21 minutes.

65skywalkingRuns60 / 100Java25,019

APM, Application Performance Monitoring System

What the test found: Maven build succeeds, all 641 unit tests pass, and the OAP server starts and serves GraphQL and admin HTTP endpoints using H2 database as the storage backend. 37 minutes.

66VictoriaLogsRuns60 / 100Go2,541

Fast and easy to use database for logs, which can efficiently handle terabytes of logs

What the test found: VictoriaLogs builds from source, starts as a daemon listening on port 9428, serves HTTP 200 on its root endpoint with correct metadata, and all 17 test suites pass cleanly with zero failures. 9 minutes.

67zoetropeRuns60 / 100Rust1,198

Watch a Claude Code or Codex session as a live flow graph, in your terminal or your browser.

What the test found: Zoetrope builds from source, passes all 216 tests, passes clippy and fmt checks, and the zoe binary reads and inspects both Claude and Codex session transcripts, outputting complete agent trees with tool counts and metadata. 4 minutes.

68openGeminiRuns57.5 / 100Go1,182

CNCF sandbox project, an open source distributed time-series database with high concurrency, high performance, and high scalability

What the test found: openGemini ts-server builds, starts, accepts InfluxDB line protocol writes on port 8086, and returns queried data through the HTTP API; the full app test suite (14 packages) passes. 48 minutes.

69micrometerRuns54 / 100Java4,911

An application observability facade for the most popular observability tools. Think SLF4J, but for observability.

What the test found: Micrometer builds and all 4682 tests pass across 39 modules; 3 pre-existing test failures remain in Dynatrace LTT percentile export, Statsd intermittent-failure timing, and Jetty12 request counting race. 30 minutes.

70greptimedbRuns50 / 100Rust6,825

The open-source observability database. One columnar engine for metrics, logs, and traces, on object storage.

What the test found: GreptimeDB 1.3.0-alpha.1 built from source under Rust 1.96.1 starts successfully in standalone mode, binds HTTP on :4000, MySQL on :4002, Postgres on :4003, and gRPC on :4001, and correctly handles CREATE TABLE, INSERT, and SELECT SQL queries via the HTTP API. 82 minutes.

71persesRuns50 / 100Go2,550

The CNCF sandbox for observability visualisation. Supports Prometheus, Tempo, Loki, Pyroscope, and more!

What the test found: The Perses API server (81M binary) launches on port 8080, health endpoint returns HTTP 200, native auth login returns a JWT token, API returns 3 projects and 17 dashboards from provisioning data, the Perses CLI (37M binary) responds to all commands, 24 Go test suites and 162 UI tests all pass. 20 minutes.

72khiRuns50 / 100Go2,086

A log viewer for Kubernetes troubleshooting

What the test found: make setup completed successfully: npm install, protobuf codegen, backend codegen, MSDF font atlas, Angular production build, and Go compiler all built without errors; all 1417 frontend tests pass; all backend Go tests pass (skipping Cloud Logging tests due to missing GCP credentials); the khi binary runs and serves... 14 minutes.

73logbookRuns50 / 100Java2,070

An extensible Java library for HTTP request and response logging

What the test found: Logbook v4.3.0-SNAPSHOT builds and all tests pass with JDK 17.0.8.1 (Adoptium Temurin) downloaded from GitHub releases due to no system JDK being available on the Ubuntu 24.04 container. 20 minutes.

74langsmith-sdkRuns50 / 100Python1,081

This repo was archived on Oct 16, 2026. Python SDK moved to langchain-ai/langsmith-python and JS SDK to langchain-ai/langsmith-javascript.

What the test found: Python SDK installs in a venv and all 2553 unit tests pass; one source file fixed for httpx2/pytest-httpx compatibility. JS SDK installs but tests fail on Node 18 (crypto.randomUUID missing) and vitest shows 113 pre-existing failures. 45 minutes.

75nightingaleRuns44 / 100Go13,512

Open-source, self-hosted alerting for metrics, logs, and databases. Includes a built-in AI agent and MCP server.

What the test found: n9e server builds, starts on port 17000 with SQLite+embedded TSDB, serves SPA and REST API, authenticates root user login, responds to Prometheus queries, and exposes a built-in MCP endpoint with 30+ tools over Streamable HTTP transport. 13 minutes.

76mcp-context-forgeRuns33.3 / 100Python4,875

An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails...

What the test found: ContextForge gateway installed in .venv, server launches and responds on HTTP, 592+ unit tests pass against real SQLite and auth infrastructure. 32 minutes.

77open-multi-agentRuns with mocks92 / 100TypeScript7,070

Self-hosted TypeScript agent runtime with durable approvals and verifiable run records. Own it, approve it, audit it.

What the test found: The monorepo installs, builds, passes 2350 unit tests across all 4 workspaces plus the benchmark harness, and the CLI prints help, all with Node.js 22, no API keys or network required. 4 minutes.

78kube-state-metricsRuns with mocks92 / 100Go6,237

Add-on agent to generate and expose cluster-level metrics.

What the test found: Binary builds (87MB), all 17 unit test packages pass with -race, server starts and serves HTTP 200 on /healthz, /livez, /readyz, self-metrics (Go runtime) on port 8081, but requires a real Kubernetes API server to produce kube_ metrics. 12 minutes.

79odigosRuns with mocks92 / 100Go3,672

Distributed tracing without code changes. 🚀 Instantly monitor any application using OpenTelemetry and eBPF

What the test found: The Odigos monorepo Go modules common, api, and status each pass their test suites. The CLI binary compiles and runs. The frontend webapp builds as a static Next.js site. The autoscaler, k8sutils, and collector configgrpc module tests cannot execute due to missing kubebuilder/etcd infrastructure or build toolchain... 28 minutes.

80MyPerf4JRuns with mocks92 / 100Java3,566

High performance Java APM. Powered by ASM. Try it. Test it. If you feel its better, use it.

What the test found: The MyPerf4J Java project builds with Maven and all 75 unit tests pass (62 in MyPerf4J-Base, 8 in MyPerf4J-Core, 5 in MyPerf4J-ASM). The ILoggerTest was fixed to initialize config properties, and a mock InfluxDB v2 server was started for the InfluxDbV2ClientTest. 18 minutes.

81subtraceRuns with mocks92 / 100Go3,101

Network inspector for your backend

What the test found: subtrace builds from source with Go 1.24.2, passes its fd package unit test, and the binary responds correctly to version and proxy commands. 5 minutes.

82future-agiRuns with mocks92 / 100Python2,286

Open-source, end-to-end platform for evaluating, observing, and improving LLM and AI agent applications. Tracing · Evals · Simulations · Datasets · Gateway ·...

What the test found: Frontend dependencies installed with yarn, unit and integration tests pass (11/11) using jsdom environment. Backend dependencies installed via uv sync (441 packages). Django 5.1.8 and Django REST framework import correctly. Backend cannot fully start or run tests without Docker for PostgreSQL, ClickHouse, Redis, and... 50 minutes.

83GoModelRuns with mocks92 / 100Go1,322

AI gateway / AI control plane / AI proxy written in Go. Unified OpenAI-compatible and Anthropic-compatible API for OpenAI, Anthropic, Gemini, Groq, xAI...

What the test found: The GoModel binary builds and runs, serving health (200 OK), dashboard (200 OK), and models endpoint (200 OK); all 101 test packages (internal, e2e, contract) pass with zero failures. 10 minutes.

84kindlingRuns with mocks92 / 100Go1,062

eBPF-based Cloud Native Monitoring Tool

What the test found: The kindling-collector Go binary builds and runs in the container, serving Prometheus metrics on port 9500 (HTTP 200) and initializing all analyzers (network, tcpmetric, tcpconnect, cpu, k8sinfo), with the cgo receiver linked against a stub libkindling.so because the real C++ eBPF probe cannot be compiled without Qt5... 17 minutes.

85kubesharkRuns with mocks86 / 100Go12,143

eBPF-powered network observability for Kubernetes. Indexes L4/L7 traffic with full K8s context, decrypts TLS without keys. Queryable by AI agents via MCP and...

What the test found: Go toolchain installed, project builds, binary produces valid output for help and version commands, and all 4 test packages (cmd, config, kubernetes, semver) pass with every subtask succeeding. 10 minutes.

86robustaRuns with mocks56 / 100Python3,142

Better Prometheus alerts for Kubernetes - smart grouping, AI enrichment, and automatic remediation

What the test found: All 422 unit tests pass on Python 3.12, the Flask web app loads and accepts connections on port 9876 returning 500 on /healthz (expected: healthz requires Web.init which needs a Kubernetes cluster), and 3 infrastructure-dependent tests (k8s_utils, kind_cluster) are skipped as expected without a Kubernetes cluster. 12 minutes.

87calicoCould not verify71.1 / 100Go7,427

Cloud native networking and network security

What the test found: Go 1.27.1 installed, 38 test packages pass, 6 Calico binaries (calico-node, calicoctl, calico-ipam, kube-controllers, confd, apiserver) build and execute, libcalico-go library compiles and passes vet; felix and cni-plugin install/integration tests fail without system-level C headers or a Kubernetes cluster. 30 minutes.

88ecaptureCould not verify65 / 100C15,580

Capture SSL/TLS plaintext with eBPF, no MITM proxy or custom CA installation. Supports Linux and Android on x86_64\arm64 and loong64.

What the test found: eCapture binary (bin/ecapture) built as a 41 MB statically-linked ELF on Linux x86_64, runs and prints help/version output, 21 test packages pass, but real probe capture requires root and was not verified. 26 minutes.

89beylaCould not verify20 / 100Go2,153

eBPF-based autoinstrumentation of web applications and network metrics

What the test found: Prebuilt Beyla v3.36.0 binary launches and prints version. 11 Go test packages pass. Example HTTP service compiles. Source build fails due to missing eBPF bytecode generation requiring Docker or clang >=19. 44 minutes.

90pupCould not verify20 / 100Rust1,047

Give your AI agent a Pup, a CLI companion with 200+ commands across 33+ Datadog products.

What the test found: could not be verified; the log shows where it stopped. 34 minutes.

91trinityCould not verify20 / 100Python794

Self-hosted AI Agents Platform for AI Native companies. Claude Code, Codex, Gemini, OpenCode agents. Apache 2.0.

What the test found: could not be verified; the log shows where it stopped. 29 minutes.

92TracekitCould not verify10 / 100Python841

Tracekit records what your AI agents do, decides each tool call before it runs, and gives you a signed record anyone can verify offline: tamper-evident...

What the test found: could not be verified; the log shows where it stopped. 36 minutes.

93coze-loopCould not verify0 / 100Go5,800

Next-generation AI Agent Optimization Platform: Cozeloop addresses challenges in AI agent development by providing full-lifecycle management capabilities from...

What the test found: could not be verified; the log shows where it stopped. 87 minutes.

-loongcollectorNot yet verified-C++2,218-
-ClickStackNot yet tested-TypeScript10,064-
-alloyNot yet tested-Go3,702-
-agent-consoleNot yet tested-JavaScript800-
-cloudproberNot yet tested-Go722-
-telescopeNot yet tested-Python715-

runs installed and started with its real dependencies. runs with mocks started after stand-ins replaced external services such as a database or a third-party API. could not verify neither the standard agent nor the stronger one got it running within the time limit; the log shows where it stopped.

How we tested

On this list as of the latest test: 76 projects ran as-is, 10 with mocks, 7 could not be verified, 6 still waiting. Languages tested: C, C++, Go, Java, JavaScript, Python, Rust, Svelte, TypeScript. Every attempt used a clean single-use machine, the subject at a pinned version, and a 90-minute limit for standard projects; the complete procedure is on the methodology page.

Frequently asked questions (FAQs)

How is this list ranked?

By measurement, not opinion: projects Argusic installed and launched on a fresh machine come first, then those that ran with mocks in place of external services, then those it could not verify. Ties go to the Argusic Score, then how popular it is on its own source.

Why are some projects unranked?

6 projects are still waiting for a test or for a finished attempt. They are listed without a rank until Argusic has measured them.

Where is the evidence?

Every row links to the project's Argusic page, where each run has a full log and a terminal recording stored with a sha256 fingerprint. The same pages exist for every one of the tested projects, on this list or not.

More lists in this category

All lists: Best. All tested projects: subjects.