Argusic

Argusic agents test real software end to end: install, launch, debug, verify. As an Amazon Associate I earn from qualifying purchases.

[Argusic @argusiccom]

Argusic: Does it run?

Argusic is an independent project that tests real software end to end. Our agents install each project on a clean machine, launch it, debug it and verify that it works. Every verdict links to its own log and recording, and nothing is published until a person approves it.

🧪 Subjects
Every project we have installed and run, each with its verdict and the log and recording behind it.
https://t.co/z3D4gn9ffG

☁️ Servers
GPU cloud and VPS prices with a verification state on every row, plus the spin-up times and billed rates we measure on our own benchmark fleet.
https://t.co/kdyDnilg7E

💻 Hardware
The best laptops and desktops for programming, AI and data science. Filter by RAM, CPU, screen, weight, OS and the AI model each machine can run, with prices updated hourly.
https://t.co/lDlGfWYWWi

Follow us for new verdicts, price changes and hardware picks, and see everything at https://t.co/9LTsVwt5E7

As an Amazon Associate I earn from qualifying purchases.

#AD #Argusic

[Argusic @argusiccom]

Walrus: 🔥 Fast, Secure and Reliable System Backup, Set up in Minutes.
Clivern/Walrus · Go · MIT · 546 stars

Walrus tested: 1 error, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit c24b090.
Install to first launch: 2 min
Whole session: 6 min
Errors met: 1, fixed without help: 1
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "Walrus builds from source, all 19 unit tests pass, the tower binary starts on port 8000 and responds to HTTP health checks with 200 OK."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Walrus #Backup #Clivern #DigitaloceanSpaces #Hacktoberfest #HighAvailability #Minio #Argusic

[Argusic @argusiccom]

dsh-univer-office: Give DeepSeek Harness a real office environment.
dream-num/dsh-univer-office · TypeScript · Apache-2.0 · 562 stars

dsh-univer-office tested: it runs, but only against stand-in services, never the real thing.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 5ce30f6.
Whole session: 22 min
Errors met: 5, fixed without help: 5
Argusic Score: 92/100
Depth: checked against stand-in services only

The agent's own closing words: "The repository installs with the project-pinned Node 22 and pnpm 11.23.0, builds lib/ and artifacts/, passes lint, typecheck, format:check and the entire test suite (10 suites green, 99 comparison tests), and the built Univer Viewer boots and renders its UI in headless Chrome in this container."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#DshUniverOffice #DeepseekHarness #DeepseekHarnessPlugin #Dsh #DshPlugin #Office #OfficeHarness #Argusic

[Argusic @argusiccom]

autopulse: 💫 automated lightweight service that updates media servers like Plex and Jellyfin based on notifications from media organizers like Sonarr and Radarr.
dan-online/autopulse · Rust · MIT · 553 stars

autopulse tested: nothing broke, ran first time in 6 minutes.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 6c98b6b.
Install to first launch: 6 min
Whole session: 11 min
Errors met: 0, fixed without help: 0
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "autopulse v2.0.0 builds from source with all vendored dependencies, all 192 tests pass, and the HTTP server starts and correctly processes API requests with SQLite on localhost:2875."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Autopulse #Autoscan #Emby #Jellyfin #Plex #Servarr #Rust #Argusic

[Argusic @argusiccom]

giselle: Giselle: AI App Builder.
giselles-ai/giselle · TypeScript · Apache-2.0 · 558 stars

giselle tested: it runs, but only against stand-in services, never the real thing.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 7dca224.
Install to first launch: 1 min
Whole session: 11 min
Errors met: 5, fixed without help: 4
Argusic Score: 88/100
Depth: checked against stand-in services only

The agent's own closing words: "Giselle monorepo installs, all SDK packages build, 653 tests pass, types check, and the Next.js dev server starts and responds with HTTP 307 at localhost:3456."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Giselle #Agent #AgentBuilder #AgenticAi #Ai #AiAgent #AiAgents #Argusic

[Argusic @argusiccom]

smart-ralph: Spec-driven development with smart compaction.
tzachbon/smart-ralph · Shell · MIT · 559 stars

smart-ralph tested: 1 error, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit ac7251a.
Install to first launch: 1 min
Whole session: 23 min
Errors met: 1, fixed without help: 1
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "Smart Ralph plugin suite fully passing: 342 Bats tests and 13 Python tests all green, with stop-watcher, state management, integration, prototype phases, interview framework, phase gates, workflow guardrails, and platform validation all verified"

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#SmartRalph #AgenticCoding #Ai #AiAgents #AutonomousCoding #Claude #ClaudeCode #Argusic

[Argusic @argusiccom]

ComfyUI-OmniVoice-TTS: OmniVoice TTS nodes for ComfyUI - Zero-shot multilingual text-to-speech with voice cloning, voice design, and multi-speaker dialogue.
Saganaki22/ComfyUI-OmniVoice-TTS · Python · Apache-2.0 · 560 stars

ComfyUI-OmniVoice-TTS tested: 7 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 3fe8e42.
Whole session: 5 min
Errors met: 7, fixed without help: 7
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "5 OmniVoice TTS ComfyUI nodes (Longform TTS, Voice Clone, Voice Design, Multi-Speaker, Whisper Loader) installed and registered; OmniVoice-bf16 model loads from HuggingFace cache, generates valid 24kHz audio waveform from text on CPU in ~9 seconds for 2-second utterance"

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#ComfyUIOmniVoiceTTS #Comfyui #CustomNodes #TextToSpeech #Tts #VoiceClone #VoiceCloning #Argusic

[Argusic @argusiccom]

dubbo-go-pixiu: Based on the proxy gateway service of dubbo-go, it solves the problem that the external protocol calls the internal Dubbo cluster.
apache/dubbo-go-pixiu · Go · Apache-2.0 · 565 stars

dubbo-go-pixiu tested: nothing broke, ran first time in 2 minutes.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 54e6c4c.
Install to first launch: 2 min
Whole session: 10 min
Errors met: 0, fixed without help: 0
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "Go 1.25.0 installed from source, project compiles to a 119MB binary, all 71 test modules pass, and the gateway binary starts and answers HTTP on port 8888 returning the expected 502 when no backend is connected."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#DubboGoPixiu #ApiGateway #ApiRest #CloudNative #Cloudnative #Dubbo #DubboCluster #Argusic

[Argusic @argusiccom]

vaahcms: VaahCMS is a laravel based open-source web application development platform shipped with a headless content management system (CMS).
webreinvent/vaahcms · PHP · MIT · 573 stars

vaahcms tested: 5 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 8d7898f.
Install to first launch: 20 min
Whole session: 21 min
Errors met: 5, fixed without help: 5
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "VaahCMS package installed in a fresh Laravel project on SQLite, all 29 migrations applied, dev server answers /api/health with HTTP 200 and body OK, and the default Laravel test suite passes 2/2 tests."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Vaahcms #Admin #Application #Backend #Buefy #Bulma #Cms #Argusic

[Argusic @argusiccom]

para: Multitenant backend server for building web and mobile apps rapidly.
Erudika/para · Java · Apache-2.0 · 574 stars

para tested: 2 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit dc9522c.
Install to first launch: 1 min
Whole session: 19 min
Errors met: 2, fixed without help: 2
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "Para 1.55.4-SNAPSHOT builds from source, starts an embedded Jetty server on port 8080 responding HTTP 200, and passes all 172 unit tests and 15/15 end-to-end REST API integration tests with Lucene search and H2 persistence via MockDAO."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Para #Api #ApiServer #Baas #BackEnd #Backend #BackendApi #Argusic

[Argusic @argusiccom]

simple-todo-c: A lightweight, native Windows Todo application built with pure C and the Win32 API.
Efeckc17/simple-todo-c · C · MIT · 574 stars

simple-todo-c tested: 4 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit bcdf3d5.
Install to first launch: 22 min
Whole session: 23 min
Errors met: 4, fixed without help: 4
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "The C/WinAPI todo app cross-compiles to a 60KB exe and launches under Wine, showing a 'Todo Application' window on the virtual display, confirming the GUI is functional."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#SimpleTodoC #Gui #Minimal #Native #Todo #TodoApp #TrayApplication #Argusic

[Argusic @argusiccom]

otari: Open-source, OpenAI-compatible LLM gateway you run yourself.
mozilla-ai/otari · Python · Apache-2.0 · 624 stars

otari tested: 2 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 1fbdc65.
Install to first launch: 1 min
Whole session: 20 min
Errors met: 2, fixed without help: 2
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "The Otari LLM gateway installed from source with uv, boots as a standalone server on port 8000 serving the management API, bootstrap, health, docs, OpenAPI spec, and tutorial root page, and the OSS edition smoke test exercises all key lifecycle stages (health, bootstrap, user/keys, providers, routing, completions, usage, mail) against a mock provider."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Otari #Ai #AiGateway #Anthropic #ApiKeyManagement #Budgets #CostTracking #Argusic

[Argusic @argusiccom]

tuliprox: Flexible IPTV playlist processor & proxy in Rust 🦀 - supports M3U, Xtream Codes, HDHomeRun, Plex, Emby & Jellyfin.
euzu/tuliprox · Rust · MIT · 580 stars

tuliprox tested: 5 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit cfc78fd.
Whole session: 20 min
Errors met: 5, fixed without help: 5
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "tuliprox builds and passes all 5532 unit tests; the binary starts in server mode, serves /healthcheck on port 8901 returning HTTP 200, and processes real M3U playlists in CLI mode."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Tuliprox #CatchUp #Emby #Iptv #IptvProxy #Jellyfin #Jthreadfin #Argusic

[Argusic @argusiccom]

Argusic daily report, 17 October 2026 (UTC)

On 17 October an AI agent tried to install and run 95 open-source projects on clean machines, with nobody helping: 114 attempts in all, 7 of them cut short by our own tooling and not counted.

Runs: 63
Runs with mocks: 21
Could not verify: 7
Not yet verified: 4

Five from the day:
• SurfSense (16,497 stars): Runs
https://t.co/hvVFNK3XuH
• d2r-mobile (3,696 stars): Not yet verified
https://t.co/x1wutw8rLP
• deckcraft (3,686 stars): Could not verify
https://t.co/sEghlqyFp0
• scroll-craft (3,585 stars): Runs
https://t.co/bZwBa9Bq0J
• agentty (618 stars): Could not verify
https://t.co/3OqrviXznb

All 95 projects, each with its log and recording:
https://t.co/RLcOS0mEGV

How it is measured: https://t.co/p4x5wxPWOV
Every project tested so far: https://t.co/z3D4gn9ffG
GPU cloud and VPS prices: https://t.co/kdyDnilg7E
Laptops and desktops for programming and AI: https://t.co/lDlGfWYWWi

[Argusic @argusiccom]

Adrian: Open-source runtime AI agent security tool - monitors and controls AI agents, catching malicious tool use, prompt injection, and policy drift in real time, before the agent acts.
secureagentics/Adrian · Python · Apache-2.0 · 585 stars

Adrian tested: 6 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 73c6c1c.
Whole session: 18 min
Errors met: 6, fixed without help: 6
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "Python SDK v1.2.1 installed and all 422 tests pass; Go backend binary compiled and running with SQLite database seeded, responding on health/ready endpoints (200) and all REST API routes; Next.js frontend dashboard built as standalone and starts successfully."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#Adrian #Aarm #AgentSecurity #AgenticAi #AgenticSecurity #Agents #Ai #Argusic

[Argusic @argusiccom]

laravel-top: 🚀 Real-time monitoring from the command line for Laravel applications.
leventcz/laravel-top · PHP · MIT · 583 stars

laravel-top tested: it runs, but only against stand-in services, never the real thing.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 1f2ba3c.
Install to first launch: 10 min
Whole session: 20 min
Errors met: 3, fixed without help: 3
Argusic Score: 92/100
Depth: checked against stand-in services only

The agent's own closing words: "The laravel-top package installs with Composer under PHP 8.4.23 and its own test suite passes (19 tests, 33 assertions); phpstan and pint report 13 pre-existing style/analysis issues each and the PHPStan turbo accelerator warning shows the static build cannot load dynamic extensions."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#LaravelTop #Apm #Cli #CommandLineTool #LaravelPackage #Metrics #RealtimeMonitoring #Argusic

[Argusic @argusiccom]

gemini-web-to-api: ✨Reverse-engineered API for Gemini web app.
ntthanh2603/gemini-web-to-api · Go · MIT · 592 stars

gemini-web-to-api tested: it runs, but only against stand-in services, never the real thing.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 3633170.
Install to first launch: 2 min
Whole session: 18 min
Errors met: 0, fixed without help: 0
Argusic Score: 92/100
Depth: checked against stand-in services only

The agent's own closing words: "Go 1.27.2 installed, project builds, server launches and serves /health on port 4981 responding HTTP 200, all 9 test suites pass (85+ subtests), Gemini auth fails with dummy cookies as expected"

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#GeminiWebToApi #Ai #Api #Chatbot #Claude #Docker #FiberFramework #Argusic

[Argusic @argusiccom]

next-plaid: NextPlaid, ColGREP: Multi-vector search, from database to coding agents.
lightonai/next-plaid · Rust · Apache-2.0 · 595 stars

next-plaid tested: 1 error, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 00e26aa.
Whole session: 14 min
Errors met: 1, fixed without help: 1
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "The Rust workspace compiles, all 1062 unit tests pass, both binaries respond to help/version, NextPlaid API serves CRUD endpoints (index create/add documents/search returning real scores), and ColGREP builds indexes and performs semantic code search with ONNX model inference."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#NextPlaid #AgenticRag #Cli #Grep #MultiVector #VectorDatabase #Rust #Argusic

[Argusic @argusiccom]

agent-manager: The fastest developer workflow for every AI coding agent.
YoanWai/agent-manager · Go · Apache-2.0 · 601 stars

agent-manager tested: 2 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit ecd836f.
Install to first launch: 12 min
Whole session: 15 min
Errors met: 2, fixed without help: 2
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "agent-manager binary builds from source, all 28 test suites pass, formatting and vet checks are clean, and the binary prints its version and help text successfully"

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#AgentManager #AiAgents #Bubbletea #ClaudeCode #Cli #CodeReview #Codex #Argusic

[Argusic @argusiccom]

reactive-interaction-gateway: Create low-latency, interactive user experiences for stateless microservices.
Accenture/reactive-interaction-gateway · Elixir · Apache-2.0 · 602 stars

reactive-interaction-gateway tested: 7 errors, no human help, still ran.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit b65435c.
Whole session: 24 min
Errors met: 7, fixed without help: 7
Argusic Score: 100/100
Depth: checked against the real service

The agent's own closing words: "Erlang OTP 25 + Elixir 1.18.5 installed from Ubuntu packages and source build. RIG project compiles in dev/test/prod. Test suite runs 273 tests with 226 passing. pg2-to-pg migration applied. Project was designed for Elixir 1.11.3/OTP 23 and has partial compatibility with OTP 25/Elixir 1.18."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#ReactiveInteractionGateway #ApiGateway #CloudeventsSchema #EventDrivenArchitecture #EventDrivenMicroservices #EventSubscriptions #Frontend #Argusic

[Argusic @argusiccom]

gitlab-mcp: MCP server for GitLab: 100+ tools for projects, MRs, issues, pipelines, wikis and more via the GitLab REST/GraphQL API.
zereight/gitlab-mcp · TypeScript · MIT · 2,047 stars

gitlab-mcp tested: it runs, but only against stand-in services, never the real thing.

It was installed and run by an Argusic agent on a clean machine, with no human help, from the repository pinned at commit 95edb49.
Install to first launch: 20 min
Whole session: 22 min
Errors met: 1, fixed without help: 1
Argusic Score: 92/100
Depth: checked against stand-in services only

The agent's own closing words: "The @zereight/mcp-gitlab MCP server version 2.3.2 installs, builds, and passes its full mock test suite (all tests pass) using a built-in mock GitLab server that replaces the real GitLab API."

The whole session is in this video, unedited, at the speed it actually happened. Every number above is read from that run's own log.

Follow for a new repository put through this every day.

#GitlabMcp #Ai #Claude #Cursor #Devtools #Gitlab #GitlabApi #Argusic