#Observability
200
2026-09-03
Show HN: Aura – a Rust agent that investigates and fixes production incidents
S
We run a SaaS that handles petabytes of data. Our SRE team experimented with using claude, openclaw, langchain, etc. within our incident response workflows. We struggled with overflowing context, lethal trifecta vectors, hallucinations, and burned a lot of frontier tokens mostly on easy work. Approval fatigue was a challenge, and we drew a hard line at relaxing permissions in production. Long story short, we built and open-sourced AURA, a Rust-based harness specifically designed for the type of operations work which routinely involves large volumes of telemetry data and coordinated investigations across many domains of knowledge / state. We have found even on open-weights models, root cause accuracy has been very good, and remediation actions are guarded with human-in-the-loop. AURA runs from a centralized configuration file where workers are defined and scoped to task domains, (e.g. logging review, metrics analysis, and git/scm related queries). All permissions, tools access, LLM backend(s), worker prompts, and the main coordinator prompt are also defined in code. Permitted tool use is enforced deterministically outside the agent's context, so the agent cannot grant itself capabilities through prompting. AURA also handles the rest of the execution layer, such as human approval for sensitive actions and fault tolerance. A lot of thought went into context window management: large tool outputs and worker responses are persisted to disk and agents are given tools to slice/read as needed. The project is Apache 2.0. You can install rpm/deb/brew natively or pull the source and build your own. It runs either as an assistant on your workstation, or as a daemon that you can connect to via any OpenAI-compatible frontend. The GitHub repo is here: https://github.com/mezmo/aura/ The whole thing is free/Free. No signups or open core trapdoor. Demo: We put together an 8-minute video showing a checkout outage resolved by following evidence from several different systems: https://www.youtube.com/watch?v=TERHoRzT8cE tl;dw: 502s from a checkout service cause Grafana to trigger an incident via PagerDuty. AURA (using DeepSeek-V4-Flash) uses its workers and tools to correctly determine root cause as a memory leak in a downstream service, and attributes the incident to a defect in a recently merged PR. The demo then concludes with a human-gated tool call to GitHub, where AURA documents the exact lines of code that are causing the problem, and recommends a fix. Once the PR is deployed, AURA validates that the errors are clearing and transactions are no longer failing. The whole time, AURA’s leaving an audit trail of OTEL events into Phoenix. Technical details: AURA uses an agent coordinator that drafts, executes, and supervises DAG flows through user-defined worker agents to solve complex investigations in parallel. Worker agents are bound to strongly typed durable artifacts used as evidence packets, and the system discourages wasteful recalculation of remote data. AURA also has a powerful header routing system that doesn’t allow the agent to actually touch your keys. HITL interrupts over webhook are first class citizens and have a schema that is easy to adapt to your own workflow. HA deployment options are coming soon. What’s still rocky: - We’re currently working on the async input system for when you’re running AURA as a service. Right now the API just accepts a request, and streams messages until the main loop completes. It’s easy to integrate into a workflow, but the logic required winds up being heavier than it ought to be. - Because there’s no inbound webhook interrupt mechanism, automating AURA for IR requires middleware to invoke it or it needs to poll an MCP for alert escalations. We’re looking for users, contributors, guidance on where to take this next, or just GitHub stars. Looking forward to reading and responding to feedback here.
2026-09-11
Show HN: Open-source simulation testing infra for voice agents
A
Hey HN, we’re Nischal & Naman. We’re brothers, and together we’re building an open-source platform for simulation based testing of voice agents (try it out in 5 mins - https://docs.egma.ai/docs/get-started/quickstart, 2 min demo video - https://youtu.be/wgDWEe5UAUY) Platforms that help you do simulation testing already exist. But they all charge a heavy premium on top of inference costs. We believe if the industry truly wants to scale simulation testing of voice agents, we need to stop charging a premium on inference and start providing infrastructure to scale simulations. The project is built in a way that allows you to bring your own STT, LLM, and TTS provider keys while also providing a way for us to provide inference directly within the platform at a 0% markup so you don’t have to wrestle with multiple keys/ rate limits. A bit on why’re we’re building this - We started working on voice agents in early 2024 and since then, have worked on numerous voice ai systems - like screenless voice-powered hardware for kids[1], AI receptionist deployed in healthcare practices like med spas & therapy clinics and personal accountability coaches. Some of these were full-fledged startups others were just side projects. But we kept on encountering similar issues all throughout. One of the most frustrating parts was calling the agent again and again, reciting the same script to test its behavior. We also kept on encountering new issues all the time in production that we couldn't have simulated pre launch. This frustration led us to a bigger question: how can developers trust the voice agents they’re shipping? We believe building that trust takes two things - - First, you need a way to test the major scenarios your agent will face in production before you release it, without having to make every call yourself. But you can’t test for everything. Real world is too messy to predict in advance. - Which brings us to the second: you need a way to find issues once your agent is in production, whether it’s handling a few dozen conversations or millions. Detect drift in known behaviors AND surface unknown-unknown ways in which your agent is going wrong. We think it’s a really hard problem to solve. And having felt it firsthand, we’re deeply motivated to take a shot at it. Today we’re launching the simulation testing side of the platform. We feel it’s mature enough that real teams can depend on it. For eval design, we took inspiration from anthropic’s evals design[2] and extended it to voice systems. For technical & business model design we borrowed ideas from Langfuse[3]. Our stack is postgres, clickhouse & minio. Its easy to self-host[4] & the code has a permissive MIT license. We also have managed cloud version. We’ve written more about our [testing]( https://docs.egma.ai/docs/core-philosophies/testing-philosop...) and [monitoring]( https://docs.egma.ai/docs/core-philosophies/monitoring-philo...) philosophy in the docs. We’d love feedback from the HN community and people building voice agents - how are you testing today, what’s working and what’s frustrating? We’ll be in the comments. Thanks! [1] https://x.com/theBhulawat/status/1966200231705595932?s=20 [2] https://www.anthropic.com/engineering/demystifying-evals-for... [3] https://github.com/langfuse/langfuse [4] https://docs.egma.ai/self-hosting/get-started
2026-09-15
Matthew0822/ToolReplay
A
Audit AI agent tool-call transcripts: hash-chain sealing, deterministic replay, and scope overreach checks. Dependency-free Python CLI.
2026-09-12
FankChen/tracecrate
A
Local-first AI agent trace workbench. Inspect Claude Code, Codex and OTLP logs, compare runs, and export privacy-conscious reports. No backend or API keys.
2026-08-24
halarewich/slotstream
A
Polyglot real-time observability platform for the Solana blockchain
2026-08-23
dbl8005/sitrep-panel
A
A live local HTML report Claude keeps updated as it works — status board, running narrative, screenshots, newest-first log, live-reload. No cloud, no build step.
2026-10-23
keegan0521/CinderVM
A
Deterministic bytecode VM for agent execution - verified images, snapshots and a hash-chained replay journal
2026-09-21
khangzxrr/tmon
A
A full-screen terminal status monitor for home servers and NAS boxes. One screen answers "is everything OK?" and shows what's going on: CPU, memory, storage and RAID, disk SMART, UPS, Docker services, cameras, backups, upcoming timers, and a live list of the web requests hitting your reverse proxy.
2026-10-25
Charonyuu/whatsmyagentdoing
A
See what every Claude Code and Codex agent on your machine is doing — live, in a little office.
2026-09-13
matthart1983/kernwatch
A
Linux kernel observability in your terminal. Built with Rust and Ratatui.
2026-10-03
juliankang4/spark-scope
A
Read-only dashboard and rack panel for DGX Spark / ASUS GX10 / MSI EdgeXpert nodes running vLLM or SGLang
2026-10-02
flaviocopes/factorylog
B
See what your coding agents did for you: a timeline of each day and week on your Mac, and where your time went, project by project
2026-11-02
Hona/strobe
A
Measure and explain performance: web pages, Electron apps, Bun/Node, any process. Agent-native CLI + MCP.
2026-10-07
thkukuk/sd-report-collector
A
Server counterpart to "systemd-report upload". Accepts report pushes over mTLS, stores each one zstd-compressed under a per-hostname directory, and runs configured plugins against the saved file.
2026-10-29
bjarneo/shoulder
A
Watch over your agents' shoulder. A TUI that shows Claude Code, Codex, Gemini CLI and opencode work in your project, live.
2026-09-13
PG-Circuit/pg-circuit
A
Circuit breaker for production PostgreSQL. Native extension that can ALLOW, WARN, or BLOCK dangerous SQL based on live database state. Apache 2.0.
2026-09-17
donvito/agent-monitor
B
Visualize your Codex and Claude Code subagents. Advanced traces view, tokens breakdown, and cost usage is also supported.
2026-10-15
lufficc/clash-traffic
B
See where your Clash Verge traffic goes — by app, site, proxy node, subscription and device.
2026-10-04
Cherry/cloudflare-drizzle-tracing
B
Cloudflare Workers tracing spans for Drizzle ORM, for any driver
2026-10-14
tetriz-ai/agent-desk
B
A local macOS menu bar app that shows your AI coding sessions per tool and project, and the skills, rules, agents and hooks your coding agents load.
2026-10-13
tetriz-ai/session-intelligence
B
Help individuals and teams understand their prompting habits and improve how they work with AI coding agents. Get specific insights into prompt clarity, context, rework, outcomes and wasted effort, plus security flags. Bring your own LLM, no SaaS.
2026-10-05
atinfinity/cc-calendar
B
Google Calendar-style weekly view of your Claude Code sessions — local, live, read-only
2026-10-01
keshavbiswa/rails_event_viewer
B
A Rails engine to browse, search, and analyze structured events from Rails.event
2026-10-23
zoefix/agent-speed
B
Agent Speed — A lightweight, real-time token speed monitor for Codex and Claude Code. Track tokens per second (TPS), detect slowdowns, and monitor AI agent performance at a glance.
2026-10-02
olivierlambert/xentop-ng
B
A modern, btop-inspired resource monitor for the Xen hypervisor (demonstrator)
2026-10-01
flaviocopes/skillscout
B
Find the skills your coding agents load, see which agents can use each one, and count how often you use them
2026-09-27
agentscope-ai-java/sqlite-wal-doctor
B
Diagnose SQLite WAL misconfiguration — lock contention, unbounded WAL growth, checkpoint starvation.
2026-09-11
Show HN: DOOM in the kernel, or fibers in eBPF
B
Some time ago my coworker who was working on first version of Perforator ( https://github.com/yandex/perforator) kept talking about eBPF, so I got curious about its ISA and restrictions. That was around the time I read about DOOM on pregnancy test, so I thought: "what if someone run DOOM inside of Linux kernel in eBPF? Surely, with some restrictions, this should still be possible?" I started to play around with it somewhere in 2024 - first stripping DOOM to bare minimum that is needed for showcase, simplifying code and trying to pass first restrictions that I encountered - function limit, recursion, memory access and checks. I tried to do this by hand, but it was, well, tedious labor. So I changed direction and instead started to do some of the things with LLVM-passes - rewriting memory access, simplifying and rewriting loops etc. But there were just so many corners. And given absence of time and all, I forgot about this project. But couple months ago I thought: "LLMs are quite good nowadays, so why not try again this time with more hands". I tried different approaches on how memory access could be "virtualized", how loops can be made bounded, and in general - how to run unbounded logic on such a machine. When I got DOOM running, I got carried away. Approach was so "generic", so it would be a crime not to try to run more things. And now we have it - lua running on xdp hot path, llama2 working with softfloats, and even cpython, that did not fit initially into 1M verifier budget, is running there now thanks to freplace. There is ton of work to do in order to make programs run faster, to make integration easier, but working examples are already there.
2026-10-02
thedyerman/mikrotik-home-netflow-plus
B
Self-hosted NetFlow/IPFIX traffic monitor for MikroTik RouterOS 7 with a live web dashboard.
2026-09-30
aashish254/limen
B
Local, opt-in proxy for coding agents: measures what your agent re-sends, compiles the scaffolding out of the prompt, and enforces nothing until you say so.
2026-08-29
ccarvalho-eng/beam_console
B
An embeddable process flight recorder and process map for BEAM applications.
2026-08-25
Fanju6/sing-box-observability
B
Read-only observability panel for the sing-box /observability/v1 API
2026-10-04
hamzafer/claude-code-mods
C
Live view of what your Claude Code agent reads and writes, plus more mods.
2026-08-29
chrisgreg/boop_error_tracker
B
Sends ErrorTracker errors to your iPhone through Boop, the tiny self-hosted push notification inbox.
2026-10-20
tomtomsatnav/agent-orchestration-standard
B
A live, local-first dashboard for Claude Code sessions: project graph of agents, tokens, tool feed, replay, alerts. Plugin + terminal UI + MCP server. Includes the AOS-1 agent orchestration spec.
2026-10-01
yeet-src/agentcap
B
eBPF Prometheus exporter for AI-agent activity — per-agent tools, domains, ports, files, CPU & network — with a Grafana dashboard. A yeet service.
2026-09-27
agentscope-ai-java/sourcemap-fidelity
C
Check that source map mappings actually point where they claim — token-level verification, not just valid JSON.
2026-09-23
andududu/jeview
B
An unofficial local visualizer for Jev (TypeSafe): a live view of every call your code makes. Not affiliated with TypeSafe AI.
2026-08-31
jeonghayoon11/CrashLens
B
A focused Windows crash analysis utility for Application Event Logs.
2026-10-19
eliasbiondo/antirion
B
⚡ The fastest open-source AI gateway — one Rust binary with a built-in console to route, secure, budget and observe every LLM call across OpenAI, Anthropic, Gemini, Bedrock and more.
2026-09-07
evangelosvlachos96-dotcom/microservices-tracing-kafka-opentelemetry
C
Distributed microservices architecture with OpenTelemetry tracing, Kafka messaging, and auto/manual instrumentation, visualized in Uptrace.
2026-10-08
JinHo-von-Choi/anti-samcheonpo
B
A harness that detects, warns about, and blocks AI coding agents wasting time and money (over-verification, context drift, infinite loops, false completion) / A harness that catches AI coding agents wasting time and money
2026-10-03
cn0xroot/ArgusBPF
B
eBPF-based system & AI-agent activity monitor with a bilingual professional/plain-language web dashboard
2026-09-14
cn0xroot/CC-Monitor
B
Claude Code Monitor : Monitor & audit every action Claude Code takes on your computer.
2026-10-29
fatihfurkan46/SignalWeave
B
Signal weaving toolkit - combine noisy service metrics into stable, explainable health signals
2026-10-09
mewsdev/AiOperationsRoom
B
Sala de operações local para acompanhar sessões e subagentes do Claude Code.
2026-10-07
tsale/jevline
B
Jevline: start from one confirmed-malicious process; Jev links the rest of the incident into a timeline, execution chain and evidence table.
2026-10-23
agentshield-dev/agentshield
B
Agentshield is open source defense for Kubernetes and AWS. It learns how your workloads really talk, flags lateral movement with evidence, and proposes fixes you approve.
2026-09-21
using-system/otelyssey
B
OpenTelemetry plugins for every coding agent, in the Agent Plugins format — a marketplace that runs itself: open an issue, the repository validates, reviews, lists and follows your plugin.
2026-09-17
zidnyzd/uptime-pulse
B
Ultra-lightweight self-hosted uptime monitoring system in pure Rust (Axum, Tokio, SQLite WAL, Rustls). Fast, low memory, and multi-arch ARM64/AMD64 ready.
2026-10-07
priyankub/crowdsec_grafana_dash
C
Grafana dashboard for CrowdSec: attack map, scenarios, top IPs/paths, agents, bouncers and a live detection-to-enforcement flow. Reads the CrowdSec MySQL/MariaDB database directly.
2026-09-08
telemetry-dev/telemetry
C
Trace AI applications and agents, and monitor token usage, cost, latency, and errors with OpenTelemetry.
2026-10-03
ShadowSafin/Synapass
C
AI inference gateway and control plane for multi-provider routing, policies, caching, tools, and observability.
2026-10-01
kyroco/vitalaize
C
A wall screen for Claude Code and Codex sessions and your GitHub builds
2026-09-20
needtobreathe/AuditHook
C
High-throughput HTTP event ingestion gateway with live SSE telemetry, payload auditing, and localhost dispatch capabilities.
2026-10-03
1942853632/log-sentinel
C
Local browser extension for incident triage, log privacy and observability checks
2026-09-17
eliasstravik/herdr-agent-progress
C
Agent-reported task progress and activity for the Herdr sidebar
2026-10-02
badfisheryhy/badfisher-ai-log
C
Java application log analysis and AI anomaly diagnosis service
2026-09-15
Show HN: Pushie – Get notified with a simple webhook
D
Pushie lets you send notifications to yourself from anywhere. Create an account, claim a username, and start pushing to every device you are signed in on with a single HTTP request. No app or SDK to install. No complex authentication. Just a HTTP POST request with your message.
2026-09-19
betta-tech/jornadas-cloud-2026
D
Proyecto de ejemplo de las Jornadas Cloud 2026: de una web estática en S3 + CloudFront a un backend serverless con Lambda, API Gateway, DynamoDB y observabilidad en CloudWatch.
2026-10-05
alphaparkinc/genpark-agent-runtime-self-healing-diagnostics-skill
D
Universal Autonomous Agent Runtime Self-Healing & Diagnostics Engine. Detects configuration drift, corrupted execution state, and missing runtime resources, executing automated root-cause classification and snapshot rollback.
2026-10-29
em4nu3i69dll/service-status-api
D
Telemetria cifrada con TypeScript, agente Go, analisis Python y Ollama, Cloudflare Tunnel y lectura mTLS.
2026-09-30
Anirudh10-png/model-guard
D
Model Guardian is an AI-powered ML model monitoring platform that detects data drift, model performance degradation, and data-quality issues. It uses KS/PSI-based drift detection, performance metrics, anomaly detection, and lightweight RAG to analyze model health and provide intelligent diagnostic insights and recommendations.
2026-09-07
alanmosilva/sre-observability-course
D
Curso prático de Site Reliability Engineering e Observability
2026-11-07
meatycamp/helios-sentinel
D
A next-generation reliability platform for proactive monitoring, automated diagnostics, and failure prevention.
2026-08-24
yudaysurendra-gif/intelligent-alert-monitoring-using-aiops
D
intelligent-alert-monitoring-using-aiops
2026-11-09
AIAgentAndy/DeepAA
unrated
Local-first AI Agent gateway & observability for Codex, Claude Code, Dsh and more. Unified model management, call tracing and Token cost analysis.
2026-11-09
IainKay/ram-altman
unrated
Monitors your RAM and reports any "quite terrible" unused capacity directly to the CEO of OpenAI.
2026-11-09
taosher/signoz-open-dashboard
unrated
Turn SigNoz into a universal embedded dashboard — drop any live dashboard into your product, portal, or big screen with a single iframe.
2026-11-08
anomalyco/strobe
unrated
Measure and explain performance: web pages, Electron apps, Bun/Node, any process. Agent-native CLI + MCP.
2026-11-08
slawek19926/Vitals
unrated
Native open-source system monitor for macOS — Apple Silicon, SMC sensors, processes, disks, network and system health.
2026-11-08
BenjiNgei/real-time-operations-intelligence-accelerator
unrated
Real-Time Intelligence Operations Solution Accelerator: Dashboards, Anomaly Detection, Fabric Data A
2026-11-07
bicho-fyi/bicho
unrated
Self-hosted error tracking on Cloudflare Workers, with OTLP ingestion and MCP tools for investigating errors.
2026-11-07
sksarvesh007/antiglaude
unrated
Production compatibility gateway bridge & observability platform connecting Claude Code CLI with Google Antigravity & Gemini inference models via Anthropic Messages API
2026-11-03
Git-FED/MuseAi-Skills-Guide
unrated
(THIS IS THE PUBLIC VERSION) Muse-by-Meta usage audit companion designed for those who want to understand Muse’s actions during a multi-step task without pretending to have telemetry they can’t actually access.Check out Muse, your personal AI agent. Code: 5F2V31 https://muse.ai/join
2026-11-02
beenuar/AiSOC
unrated
Open-source AI Security Operations Center: alert fusion, LLM-agent triage, MITRE ATT&CK investigation, and a replayable decision ledger for every agent step. Self-hostable, runs with no API keys, MIT licensed. Ships an MCP server for Claude, Cursor and Continue.
2026-11-01
Forward-Future/device-monitor
unrated
Monitor CPU, GPU, memory and temperatures across all your devices (Windows/macOS app + LAN hub)
2026-10-29
arby18/OutlineThrob
unrated
Enabling real-time data visualization and anomaly detection with low-latency, and secure system architecture, utilizing a robust library framework.
2026-10-29
bagusprts/RoughObserve
unrated
Facilitates real-time data ingestion and processing through an event-driven, streaming architecture, establishing a scalable Observer platform.
2026-10-26
actioneer-hq/Actioneer-Pulse
unrated
Observability suite for production voice agents with evals, RSI !
2026-10-23
Show HN: whytime: time(1) that tells you where the wall-clock time went
unrated
time says a command took 3.8 seconds... whytime tells you where those seconds went: $ sudo whytime -- sh -c 'sleep 0.5; dd … oflag=dsync; curl -s -o /dev/null https://speed.cloudflare.com/…' wall 3.83s 47% blocked on network: speed.cloudflare.com:443 (1 connection) 34% fsync on ~/.cache 13% timer sleep (sleep) 4% on-CPU (curl) 2% other The lines always add up to 100%. When the command waits on a child or another thread, whytime follows whoever woke it up, so a shell waiting on curl is charged to the network wait curl was stuck on. The top line is what you'd have to make faster. It can also draw a flamegraph. Uses eBPF (Linux 5.18+)
2026-10-21
Show HN: Kubectl-skyline – View your Kubernetes cluster as a 3D city
unrated
Hi HN, Just wanted to sharing my kubectl plugin. kubectl-skyline, a plugin that renders your cluster as a 3D city. Why see cluster this way ? I think a vintage futuristic explorer is always fun interface to use. it's written as Go plugin, read-only list, secret values aren't displayed. Give it a try and let me know what you think.
2026-10-20
Show HN: DogCatTrace – Clapt was distilled from some random dog
unrated
https://github.com/xiebaiyuan/DogCatTrace
2026-10-20
Show HN: Making agent-to-agent interactions observable, in a fun way
unrated
I'm building this open source to show what's happening behind the scene when AI agents are interacting with each other. Just finished building the first scenario - agents discovery.
2026-10-19
LambdaBytes/sping
unrated
A modern, cross-platform ping with real-time diagnostics, network context, quality scoring, and multi-target monitoring.
2026-10-19
xhwxt/zcode-token-usage-statusbar
unrated
Real-time token usage status bar for the ZCode desktop client (Electron). Read-only on the local SQLite DB, fully offline. Installer, CLI and MCP included.
2026-10-18
scasella/claude-flightdeck
unrated
A Claude Code mod that puts a live agent dashboard in your terminal: context and cost, advisor timeline, every permission check, subagent cards and swimlanes.
2026-10-17
Cygnux-Labs/Tracekit
unrated
Tamper-evident tracing, policy gates and cross-checks for AI agents: see what the agent did, and why.
2026-10-17
Show HN: Set Span Status for Error Logs
unrated
I created a pypi library that sets span status to error when you log an error with ability to set ignore rules for specific loggers, templates or messages.
2026-10-16
Show HN: Wayseer – A native desktop app for browsing real-time data
unrated
As an IT professional, I felt a distinct lacking of a real-time tool for tracking down problems in complex systems. I have created Wayseer Desktop as a way to explore more effective ways to explore live systems. There's a very functional free version, and paid options that unlock additional features. I'm curious what you all think of the approach. I'm a systems engineer by trade (with some development experience), not a developer, so I have relied on AI to do the heavy lifting while I request features, report bugs and perform as much manual testing as I feasibly can. There are bound to be bugs... but it is still very young and has a (hopefully) long future in which refinements can be made.
2026-10-16
Atomburstofficial/geiger
unrated
A Geiger counter for AI agents — one read-only command that inventories every agent, MCP server, plugin, and AI extension on a machine
2026-10-16
Continuum-AI-Corp/OrcaReplay
unrated
OrcaReplay — Time travel for AI agents. Record, replay, fork, and debug any agent run with any model. Built by the OrcaRouter.ai team.
2026-10-16
DingoOz/llm-visuals
unrated
Live truecolor terminal dashboard for a locally running LLM: tok/s, TTFT, GPUs, context, MTP acceptance, layer and real expert routing (llama.cpp)
2026-10-16
JerryZLiu/AgentPlayback
unrated
Are you running too few agents, or too many? AgentPlayback visualizes your coding agents so you can see when each one ran, when it was blocked on you, and what it cost.
2026-10-16
Jorji49/streamtop
unrated
Terminal HLS, DASH, and IPTV stream monitor. Wire probes, TR 101 290, SCTE-35, Prometheus metrics. Rust CLI/TUI.
2026-10-16
LalitMaganti/buildprof
unrated
Records every process and file access in a build and shows it as an interactive timeline
2026-10-16
RaycarlLei/tool-journal
unrated
TypeScript execution journal for retryable tools: SQLite durability, lease fencing, explicit uncertainty, and reproducible crash tests.
2026-10-16
StellarHyperion/stellarhyperion-backend
unrated
Backend infrastructure for indexing, monitoring, and managing Hyperion cross-chain transfers.
2026-10-16
WizisCool/oh-my-cpa
unrated
Self-hosted control plane & usage observability console for CLIProxyAPI (CPA) — AI providers, client keys, streaming telemetry, and model cost snapshots.
2026-10-16
apimart001a/llm-gateway-comparison
unrated
LLM gateway comparison: self-hosted and managed AI API gateways compared on protocol compatibility, billing units, timeout and retry semantics, idempotency, quotas and observability.
2026-10-16
apromisedland/trustworthy-agent-simulation
unrated
Auditable LLM multi-agent town simulation with AgentScope, Mesa, replay, policy experiments, and Streamlit
2026-10-16
brandynfisher/ToolReplay
unrated
Audit AI agent tool-call transcripts: hash-chain sealing, deterministic replay, and scope overreach checks. Dependency-free Python CLI.
2026-10-16
dodoguardai/dodoguard
unrated
DodoGuard: Take AI Agents to Production. Test, validate, and operate AI Agents before and after deployment.for Dify, Coze, FastGPT, n8n, Tencent ADP, ByteDance HiAgent, and more.
2026-10-16
estelwalks/aitracker
unrated
Open-source, local-first AI workspace that tracks tokens & spend across Claude Code, Codex, Cursor and more — and turns every session into reusable Skills and long-term memory. / Open-source, local-first AI workspace: automatically tracks Token usage and spend for tools like Claude Code, Codex, and Cursor, and turns every AI session into reusable Skills and long-term memory.
2026-10-16
furkankly/zoetrope
unrated
Watch a Claude Code or Codex session as a live flow graph, in your terminal or your browser.
2026-10-16
jsdvjx/dsh-strata
unrated
Session strata for the DeepSeek Harness Web GUI: the transcript's scrollbar becomes a to-scale, colour-coded map of the whole run, with your own messages emphasised and clickable anchors.
2026-10-16
keelab/keelith
unrated
A production-oriented Go framework for distributed applications with explicit lifecycle, transport, governance, observability, and contract-first tooling.
2026-10-16
monitor-probe/monitor
unrated
A lightweight server probe written in Rust (Rust + axum + SQLite)
2026-10-16
ntailio/ntk
unrated
Fast, friendly CLI and TUI for Apache Kafka: inspect clusters, move messages, watch lag, and fix things from one binary.
2026-10-16
okpingnet/OKPing-Network-Monitor
unrated
Open-source network quality monitoring platform with multi-node Ping, TCPing, HTTP, DNS, IPv6 and Traceroute testing.
2026-10-16
openclaw/openclaw-enterprise
unrated
The Enterprise-grade Control Plane for modern agentic workloads
2026-10-16
sodiumsun/agenttrail
unrated
Local observability for AI coding agents. Agenttrail Map shows project structure and activity; Agenttrail Kitchen visualizes tasks and role contributions in 3D.
2026-10-16
wangmiaozero/pi-harness
unrated
The operational superset of Pi Coding Agent — everything Pi, plus observability, governance, recovery, evaluation and multi-agent orchestration.
2026-10-16
Show HN: Flow – activity based animated scenes for ClaudeCode, Pi and OpenCode 2
unrated
Hi HN, This is just a fun little project I made to help show at a glance how hard our terminals were cooking away or what is sitting idle. It's easy to add your own flow scenes to it and it is open source (MIT).
2026-10-15
ifixai-ai/iFixAi
unrated
Independent Auditing of AI Agents. Run by human or the agent itself, to answer the most crucial question in the AI Agent Economy. Is the agent doing what is supposed to do? With iFixAi you can have this answer in less than 120 seconds.
2026-10-15
Show HN: Evaluate and improve your agent steering
unrated
Hi HN, excited to be open sourcing two tools that we have built over past few months which helps to understand your own work with AI coding agents Agent Desk is a local macOS app which allows you to browse your sessions across all top coding agents - Claude code, cursor, codex, github copilot, Kiro, Antigravity. You can find sessions by your projects, revisit those conversations, understand and inspect your entire harness - rules, skills, hooks, subagents. https://github.com/tetriz-ai/agent-desk Session Intelligence is a go library which helps structure and analyse your session transcripts using your own LLM. It can help understand missing context, repeated corrections, if session reached its goal and where is effort being wasted. It can also surface leaked credentials and PII. https://github.com/tetriz-ai/session-intelligence Both are MIT-licensed. You can also combine both together and understanding your agent steering habits better. Excited about the community using it.
2026-10-15
Show HN: How many tokens are you wasting in your context window?
unrated
Very quickly vibe coded a tool to see how many tokens are being wasted!
2026-10-14
Show HN: Monitor Postgres Health
unrated
Hello, my name is Danny and I created dbhealth to monitor postgres. I have an SRE background and was working in data eng for a couple years. dbhealth provides verdicts on health and exposes these through: - statsd - prometheus - OTLP - Turbostats (the protocol that powers https://control.turbolytics.io/) dbhealth also provides table freshness by profiling the most recent row available. Happy to answer any questions. Thank you!
2026-10-14
Show HN: Proofsource. Find where AI leaves you out. And Fix it
unrated
Hi HN. I am Lokesh. I built Proofsource as an internal tool to help my finance ERP product , Rivane, to gain visibility into AI engines like ChatGPT, Claude, Perplexity, Google AI mode and such. But when I talked about it to my friends, they also saw value in it and it became a product of itself. And now proofsource had 5 paying customers(paying $1500pm in total) in the first week itself and more than 60 brands used it to get their visibility since we launched out of beta last week. What proofsource does is that it queries ChatGPT, Perplexity, Claude and Google AI Overviews using the questions or prompts that buyers normally ask and then it shows then which competitor shows up instead of their own brand for the questions/prompt. It also gives visibility into which sources the answers cite and other things like sentiments. Everything is shown to you along with what to fix and how each data and number is calculated to give 100% transparency into what we did and why and how you can use all this information to improve your visibility. We are still adding features based on initial user's feedback like auto fixing and publishing to their social and web platforms. Running ads for them for improving the visibility. We are just learning as we interact with each customer so feel free to share your feedback. And thank you for your time. :)
2026-10-14
Show HN: Problems – see where agents get stuck using your product
unrated
hi hn, i'm joey, the founder of problems.dev. agents are increasingly becoming the most important customers of products. but when something goes wrong, an agent doesn't file a support ticket or let anyone know what happened. problems was built to change that. agents submit reports via an sdk built into mcp servers or clis, but also take error reports on agent status pages. if you're building for agents or just want to learn more, glad to have a discussion below.
2026-10-13
Show HN: ClueCDC – Open-Source UI for Managing Debezium and Kafka CDC Pipelines
unrated
GitHub cluedata/cluecdc ★1 · Open-source platform to build, operate, and monitor Change Data Capture pipelines. · ClueCDC is an open-source tool for creating, managing, and monitoring CDC pipelines through a unified interface. The goal of ClueCDC is to make CDC easier for Data Engineers to operate, observe, and manage with fewer manual tasks. The project is now available on GitHub, and feedback from the community is very welcome.
2026-10-13
Show HN: LogLineAlign: Merge logs from different time zones into one timeline
unrated
https://loglinealign.com/ I vibe-coded an application that merges entries from multiple log files into a single, chronologically ordered view. It handles timezone offsets and can auto-detect some timestamp formats, with an option to specify the format manually. Files are processed on your device and not uploaded to a server. It's all open source: https://github.com/jftuga/loglinealign In the repo, there is a quick, 3 minute narrated explainer video as well as a CLI version. Also, lnav does the same thing in a TUI and is more capable.
2026-10-11
Show HN: iTerm2 System Monitor Widget
unrated
I didn't know that you can build snippets for iTerms's toolbelt. Now I know. And I built a small but wholesome system monitor which I was always missing as a sysop-dude: - CPU/memory usage - network traffic - open ports - file system moitor - folder size watch dogs for common cache/log/temp folders Each monitor lists the usage per process/app/service, it allows filtering and offers a lot of more details like PID, running user, etc.pp. You can pause the whole thing or each of the sections to not overload the system with unnecessary "pollings".
2026-10-11
Show HN: Tessary – Find the AI agent failures your sampled evals miss
unrated
Hey, I'm Akhil, co-founder of Tessary. An open-source agent reliability platform that monitors every production trace, detects issues that sampled evals miss, and investigates to find the root cause. We started up about a year ago building synthetic users for usability testing of B2B software, we built the agents that would personify real users and use the software but, we were never able to make them work reliably enough for long running sessions. We had built a massive eval set to tune our agents but, then the evaluation itself got prohibitively costly - like 5x costlier than actually running the agent since we needed to grade across the many narrow intents of our agent. That's where Tessary originated, we wanted to make agent reliability both cover more ground and be simultaneously less costly to run. Tessary works on 2 layers. L1 classifiers - which look at every agent trace and flag potentially erroneous ones extremely cheaply. We are talking 2-3 orders of magnitude cheaper than running the actual agent. L2 agents - which triage and find the root cause of the issue from the erroneous traces. These are run on SOTA models but, because we run them on already flagged traces, the time and $ spent finding the cause is much cheaper than running grading on even 1% of traces. In our testing on complex agents, at 1M traces, this approach was 5x cheaper than sampled evaluations and as your agent gets more reliable, the cost of reliability also goes down. we are open source and can be self-hosted so, you don't need to send your traces to us. We support existing `gen_ai` spec and ingestion over OTLP so, you can add this as a new exporter to your existing setup. We have a cloud version, you can quickly try the features with a $10 credit. We are looking to build more classifiers and enable people to build custom classifiers for their needs and would love to hear feedback on common failure modes people have built solutions or evaluation for Repo: https://github.com/tessaryai/tessary Website: https://tessary.ai Try for free: https://app.tessary.ai
2026-10-10
Show HN: Senro – WebMCP Evals and Observability
unrated
Hi everyone, When I first heard about WebMCP 6 weeks ago, I started thinking about what problems it might create. One question that came to mind was how developers would evaluate WebMCP tools and monitor them in production. That's when I decided to build Senro, a platform that helps devs test WebMCP tools using goal-oriented evals, trajectory evals, and more across LLMs, as well as monitor how those tools are used in production. Senro has a free plan (not a free trial). You can monitor up to 25K events/mo and run up to 250 evals/mo using your own API key (sorry, we're not venture-backed yet).
2026-10-07
RookStrainer/neubird-download
unrated
NeuBird — Security SaaS listing. Visit link via SOFTGIT. Third-party; rights belong to original authors.
2026-10-07
toothattorneymanor/new-relic-download
unrated
New Relic — Developer tools SaaS listing. Visit link via SOFTGIT. Third-party; rights belong to original authors.
2026-10-07
Show HN: Octri.dev – Generate customizable docs, 10 SDKs, and get an MCP server
unrated
Hi HN! Emad here, one of Octri's founders! I think a lot of us had a similar experience as software engineers, we build API's with great features then comes the annoying part of maintaining Docs and SDKs (if any). That by itself is a lot of work without mentioning what happens if you had multiple SDKs and have to publish updates on all the different package managers! I personally experienced that pain and was looking for solutions but most of them (if not all) are extremely expensive for an Indie, OSS, or even a startup and they actually aim for enterprise only where the big buck lives. So I decided to build all the tool that an API would need and make it available for everyone not only enterprises. Today, I would like to present Octri.dev, a little orange octupus that takes your API spec and generates a fully customizable docs site, SDKs in 10 languages (TypeScript, Python, Dart, Rust, Swift, PHP, Ruby, Java, Kotlin, Go), and gives you an MCP server as well so agents can interact easily with your API. I also added a new feature to the mix which is turned off by default and not included in your SDK unless you (the API owner) turn it on, it's the monitoring part that reports errors that happen for your SDK users back to you before they flood you with complaints and it pin points exactly where the issue happens. It's quite complicated to explain it in a short post so here's a page explaining it in detail: https://octri.dev/monitoring (Happy to answer any detailed questions you have regarding this below) We also are offering a free-tier for our indie community + we have an OSS program that gives Business tier completely for free to help our great open-source community! More about the program here: https://octri.dev/open-source We also have a free spec auditor (no sign up needed and it doesn't save your spec either): https://octri.dev/openapi-audit Glad to have this finally out and would appreciate you guys going in and testing it out and sharing your feedback and questions below! My email (if any needs it): [email protected]
2026-10-07
Show HN: StatLite – Lightweight dashboard for application monitoring
unrated
Single Go binary (~15 MB RAM), compact SQLite-based storage with configurable retention. Ideal for self-hosted app monitoring. No Prometheus/Grafana stack required. Shows app and DB health, charts for latency, http status codes, native support for Micrometer-based frameworks like Spring Boot/Micronaut/Quarkus. Simple app integration recipes for Django, Express, FastAPI, Go/Gin and others. Live demo: https://pvrlabs.xyz/statlite/demo.html
2026-10-06
Show HN: Rashomon – An independent execution record for AI coding agents
unrated
I have been working on Rashomon, an open-source execution recorder for AI coding agents. The basic idea is that the agent transcript is not ground truth. Rashomon keeps its own record of what happened and compares it against the agent’s account. For Claude Code, it records things like: * shell commands, exit codes, and whether they may have written files * tool calls and outcomes * test commands and their outcomes * subagent lifecycle and ordering * timestamps and parent/child relationships * keyed digests of command/tool inputs and working directories It builds a timeline from these events. It does not store prompts, responses, file contents, or tool output. For example, it now detects patterns like: `test fails → only test files change → same test passes` and: `test command fails → same command passes → no corresponding file edit` These can indicate that an agent changed a test instead of fixing the underlying issue. Rashomon currently runs as a Claude Code plugin or through the CLI. CLI: `rashomon watch` then: `rashomon report` Plugin: `/rashomon:report` There are still edge cases around cwd, git state, test-selection arguments, generated files, retries, and parallel execution. I would be interested in feedback from people building agents on what execution signals you would want Rashomon to capture or correlate. GitHub: https://github.com/altrace-dev-role/rashomon
2026-10-04
Show HN: UsageAtlas(OSS), All your AI coding usage insights. One beautiful view
unrated
Track AI coding usage across 14 providers. Token history, available cost estimates, credits, and usage limits in one free, open-source desktop app.
2026-10-03
Show HN: Spens sandboxed, observable coding agents
unrated
Spens is a tool created to standardized sandboxing (good enough) and observability around coding agents. It leverages docker, nono and mitmproxy to create a reproducible sandbox environment for agents, that captures all traffic and helps you understand what your coding agent is doing and when. Right now pi, opencode, claude and codex are supported out of the box, with node and python - but you can configure any agent or environment. It was created under the realization while working under different repos and agents, that 1. agents do weird stuff sometimes (like try and nuke a folder or poke around your HD) 2. statistics and usage collection is not really consistent (and some like Claude and Codex try and hide info) 3. reproducibility of agents across devices was challenging. Instead of setting up bespoke environments each time , we needed a way to automate it. And thus Spens was born. One note on the security aspect - I write good enough cause it is docker so unless your running something like gVisor it is possible for the agent to escape the sandbox (esp with a loose nono config) - all though in practice I've not seen this behavior. Second note - if your on a mac you need orb stack, as the run times take advantage of specific Linux security features that are not available via docker on mac (windows is fine since under the hood, docker leverages wsl)
2026-10-03
Show HN: What's Agent Doing – a Claude Code UI mod that explains each step
unrated
I built this because with latest coding models, Claude goes into deep-work mode with obscure commands so that I never know anymore what it's up to. This is a mod (a plugin using Claude Code's new function hooks) that draws one line above the prompt: - the current step, using the description Claude writes for each shell command (git bisect run npx vitest … → "Find the commit that broke fixed-amount coupons"), with a timer - a row per background agent, kept after the turn ends - ▸ expands into the turn's history, with outcomes and durations It reads the model's stream, tool calls and permission requests, and doesn't change anything Claude does. No network, no telemetry, MIT. Happy to answer questions about this mod or the mod building experience!
2026-10-03
Show HN: UseJunction – Find what your team's AI tools usage
unrated
A few months ago, we started getting into the problem of actually seeing how our team's AI usage and its cost actually translates into a business outcome. We recently rolled out a public version open source after a few months of pilot testing the product. The product is UseJunction ( https://usejunction.dev), and if your team is running multiple subscriptions/api from Cursor, Claude, Codex etc. this is for you to manage subscription spend, wastage and predictions. A bit of background, we had a team of 20 people running $200 Cursor subscriptions, and some were using a few others. That was over $4000 of spend. With this tool, we were able to get that to $2600 on our next billing cycle! That is a 35% decrease. We knew that people had to be facing this problem too, so we made it opensource for the community. All of these subscription providers leave behind telemetry on local machines, we use that to estimate usage. Soon we will be releasing a feature to attribute that spend to actual business outcomes that come from your project management tools. Currently considering actually adding a level of control on top of observability so that teams/orgs can actually govern AI tool usage.
2026-10-02
Alpha-Park/genpark-kullback-leibler-divergence-drift-monitor-skill
unrated
Forward and reverse KL divergence monitor detecting policy drift and reward hacking across distributions
2026-10-02
alphaparkinc/genpark-kullback-leibler-divergence-drift-monitor-skill
unrated
Forward and reverse KL divergence monitor detecting policy drift and reward hacking across distributions
2026-10-01
Show HN: EventSend – Event tracking and routing for solo founders and SaaS teams
unrated
I was tired of learning about failed payments by my customers and not from my system, so I built an event tracker that allowed me to have a searchable feed of the important events happening for my apps and route the ones I cared about to my phone so I could take action quickly. EventSend is an event tracker and a router, your system sends the important events you want to track: signups, failed payments, new subscriptions, deployments, a new purchase, etc. so they end up in a searchable and filterable feed that you can use to have quick insight about what's happening with your product. Then there's also the routing side of things, those important events you want to know the moment they happen, and not once you go into your Grafana Loki logs, can be routed to your phone, Slack, Discord, Telegram or any other system or service via a signed webhook. You can configure rules, alerts, quiet hours or smart rollups to have the signal you're looking for. All event's are retried with a reasonable backoff, circuit breakers and limiters maintain the prviders and system reliable. If we got a 401 or 403, your destination will be inactivated instead of retried and you'll be notified for you to update the configuration. I've been a software engineer for about a decade, and I can tell you this is not a weekend project done on with a single prompt pass, the infrastructure and tooling behind the project is solid, my expertise around distributed systems was put in practice with this one: distributed Elixir, Phoenix and Postgres provide the main stack, with some other technologies that support the event queues aside. So, it's a reliable system. I've launched several other projects, some running for years in production and I'm successfully using EventSend to track what's happening on all my products from my phone. There's also an MCP connector so you can ask stuff like "did anything failed during the night?" or "what did user cus_Qw3Sx did before it churned?" to get a user's timeline or any insights you're interested in. There's a forever free plan, in case you can to try it. Happy to answer any questions.
2026-10-01
Show HN: Kguardian – seccomp profiles and NetworkPolicies from eBPF traces
unrated
At the place where I work, we use the runtime’s default seccomp profile. My colleague and I wanted to see which syscalls our pods actually make, so we decided to capture and audit them and then generate a workload-specific list of syscalls. We ended up building a feature in our security tool that does exactly that. Blog post: https://dev.to/maheshrayas/part-2-the-seccomp-profile-nobody...
2026-10-01
Show HN: Dapr Dev Dashboard, a local tool for Dapr developers (incl workflows)
unrated
I'm Marc Duiker, Dapr community manager and developer advocate at Diagrid (a company founded by the co-creators of the Dapr OSS project). I've created a new free local dashboard for Dapr developers that provides insights into everything they need: applications, sidecars, components, Dapr workflow instances and execution history, actors, state management, pub/sub subscriptions, logs, and two wizard-style UI builders for creating component files and resiliency policies (no more writing YAML by hand!). I think it's a very useful companion when doing local Dapr development, especially when developing workflow applications. Please give it a try and let me know what you think! There's also a built-in Easter egg; try typing the Konami code! ;)
2026-09-30
Show HN: The Activity Monitor for the AI Era
unrated
Cortex turns a thousand processes into the apps you actually run, remembers the last 30 days, and shows what your Neural Engine and local models are doing. Everything stays on your Mac.
2026-09-29
Show HN
unrated
This is origintracer a low level observability tool of the modern web stack. It observes nginx, Gunicorn, asynchronous Python, Uvicorn and Django and Langchain. Studies their codebases into great depth and write probes and rules libraries in Python to be used with the open source software: its link GitHub: https://github.com/Humbulani1234/origintracer
2026-09-27
Alpha-Park/genpark-agent-action-trajectory-video-summarizer-skill
unrated
Multi-frame GUI computer-use action trajectory summarizer detecting dead clicks and UI rendering latency stalls
2026-09-27
Alpha-Park/genpark-realtime-voice-agent-latency-telemetry-skill
unrated
End-to-end conversational voice agent latency telemetry profiler tracking VAD, STT, LLM-TTFT, TTS-TTFB, and playout bottlenecks.
2026-09-26
Show HN: Pingularity – scheduled Ookla/iperf3 speedtest dashboard, outage alerts
unrated
I was having intermittent issues with my ISP after switching and was looking for something that would let me easily trend downtimes, speed tests, and latency, but the existing projects out there (speedtest-tracker, myspeed, netcheck, etc.) lacked a few things I needed. I gradually put together my own, it's simple to install and supports Linux, macOS, Windows, and Docker. Interactive demo with sample data: https://demo.pingularity.dev/ - Install from https://pingularity.dev then open http://localhost:9000/, local network accessible, password protectable. - Single-binary, no dependencies, self-hosted, no telemetry. - Scheduled Ookla and iperf speed tests with a wide array of settings. - Charts for visualizing download, upload, ping, jitter, and bufferbloat over time. Outage heatmap, live latency and DNS sampling. - Uptime tracking with alerts and notifications. Webhooks with native ntfy and heartbeat. - Manage retention, backups, or delete individual runs. - Prometheus/Grafana with importable dashboard and alert rules. - Customizable, change colours, disable stuff you don't want, remove chart tiles you don't care about. - MIT license, https://github.com/pingular/pingularity - I used Claude Code and Codex extensively throughout.
2026-09-26
Show HN: AgentLens – Replay and compare Codex runs
unrated
A replay button for AI coding agents. See what Codex likely read, ran, and changed, then compare two runs to find where they diverged.
2026-09-25
Show HN: Busbar self-hosted execution boundary for Agentic Apps
unrated
Came across this new project of Busbar for execution boundaries. looks very promising. Control where AI can go, what it may use, what authority it receives, what it may cost, and what evidence is retained—before it acts.
2026-09-24
Show HN: Making agent decisions better and traceable
unrated
Hey all, I'm building Mubit to help agents make better decisions. Mubit is an agent decision attribution and calibration platform per user at scale. Each user's interaction affects the way agents make decisions and they usually get cluttered over time. Mubit helps you see the decisions an agent made (per user), calibrate it (run time / context layer), and adapt it (model layer but think Jev like decision model farm for all your power customers). Our website is at https://mubit.ai and docs at https://docs.mubit.ai
2026-09-21
Show HN: WTF > Auto-check what your coding agent changed
unrated
wTF is my attempt to understand all the development that my agents are doing all the time without getting splitting headaches and standing a chance of reading a couple of pages in any given book by the end of day. It gives me an overview of what actually happened. It's also in early experimentation to see if it can help ground the agents in actual reality. Not there yet, though. Free, open source, no API, 100% local.
2026-09-20
Show HN: Fentaris, an open-source proxy for managing multiple MCP servers
unrated
Fentaris is an open-source proxy for running multiple MCP servers behind one controlled endpoint. It provides routing, authentication, policies, and observability for MCP deployments.
2026-09-19
Show HN: OpsGraph – evidence-first PostgreSQL investigations
unrated
Open-source, self-hosted PostgreSQL investigations with bounded read-only execution and inspectable evidence behind findings. GitHub: https://github.com/Arittra-Bag/opsgraph
2026-09-19
Show HN: vNetMap – A zero-knowledge E2E encrypted network topology mapper
unrated
I built this because I was tired of manually maintaining Excel sheets and Visio diagrams for my homelab documentation. The core idea is automated discovery via Nmap, but without sending plaintext network data to a hosted backend. It uses strict client-side Zero-Knowledge End-to-End Encryption. The Architecture: The Scanner: A lightweight Python agent running locally in a Docker container. It uses Nmap to scan the local subnet. The Encryption: The payload is encrypted locally before it ever leaves the network. The backend (FastAPI/PostgreSQL) only receives and stores encrypted blobs. The Frontend: Decryption happens entirely client-side in the browser (Angular), rendering the topology using vis-network. Technical Hurdles & Limitations I'm currently facing: Hostname Discovery is a nightmare: Across different vendors, devices handle broadcast differently (mDNS, NetBIOS, pure DNS). The agent catches what it can, but often falls back to raw IPs because reliable local resolution is incredibly inconsistent. Physical Connections: While Nmap easily discovers the devices and open ports on the subnet, mapping the exact physical connections (which switch port goes where, device types, Wi-Fi details) still requires manual linking in the UI. Links: Live App: https://app.vnetmap.com GitHub (Docs & Agent Setup): https://github.com/vNetMap/vnetmap-issues I'm curious to hear how you handle automated discovery in your homelabs, especially regarding the hostname resolution issue, or if you spot any flaws in this E2E approach.
2026-09-18
Show HN: One-shot resolve: package managers as graph schedulers (2x faster)
unrated
How one-shot resolve, critical-path download scheduling, coalesced postinstalls and built-in observability can make a package manager 2x faster
2026-09-18
Show HN: Pingularity – scheduled Ookla/iperf3 speedtest dashboard, outage alerts
unrated
I was having intermittent issues with my ISP after switching and was looking for something that would let me easily trend downtimes, speed tests, and latency, but the existing projects out there (speedtest-tracker, myspeed, netcheck, etc.) lacked a few things I needed. I gradually put together my own, it's simple to install and supports Linux, macOS, Windows, and Docker. Interactive demo with sample data: https://demo.pingularity.dev/ Install from https://pingularity.dev then open http://localhost:9000/, local network accessible, password protectable. - Single-binary, no dependencies, self-hosted, no telemetry. - Scheduled Ookla and iperf speed tests with a wide array of settings. - Charts for visualizing download, upload, ping, jitter, and bufferbloat over time. Outage heatmap, live latency and DNS sampling. - Uptime tracking with alerts and notifications. Webhooks with native ntfy and heartbeat. - Manage retention, backups, or delete individual runs. - Prometheus/Grafana with importable dashboard and alert rules. - Customizable, change colours, disable stuff you don't want, remove chart tiles you don't care about. MIT license, https://github.com/pingular/pingularity I used Claude Code and Codex extensively throughout.
2026-09-15
Show HN: Replay – Audit silent prompt cache misses in AI agent transcripts
unrated
I built Replay because AI agent bills spike without a single warning or error in your logs when a prompt cache breaks. Whether your team runs Claude Code, Cursor, or Aider, the provider just silently re-bills your entire conversation history at cold-write prices instead of cached read prices. Replay is a zero-telemetry, zero-dependency local Go CLI that forensically reads your on-disk log transcripts, replays them against reverse-engineered provider caching state-machines, and names the exact turn the cache broke, the cause, and the exact token leakage. Repo: https://github.com/RedRobotKK/Replay Core Subcommands (What it actually does): - `replay` : Computes total avoidable re-billed cost across your local transcripts. - `replay diff <transcript>` : Pinpoints the exact turn the cache sharded and why. - `replay advise <dir>` : Outputs structural prompt optimization adjustments based on your history. - `replay context <transcript>` : Ranks and isolates exactly what content is bloat-filling your context window. - `replay route <dir> --to <model>` : Simulates what switching models would cost, factoring cold-start prefixes. Why Go / Design Constraints: 1. Zero Dependency: Avoids messing with Python virtual environments or Docker runtimes for local diagnostics. 2. 100% Sandboxed: Absolutely no remote tracking, cloud accounts, or telemetry. Your prompt text never leaves your machine. 3. Mutant Testing Isolation: Keeps 76 real past defects frozen as re-runnable mutants to enforce code stability. It's completely free and source-available (BUSL 1.1). To check your local developer logs or audit the setup script before running, use: curl -fsSL https://replay.doctor | less
2026-09-15
Show HN: Vigilator – human-in-the-loop layer for AI agents
unrated
Hi - I'm Noah, a student working out of the UK & I've been building Vigilator for the past couple of months, Vigilator provides an agnostic solution for handling interrupts and observing agents as they work. I was inspired to build Vigilator after I realised as teams roll out AI agents across organisations there's no consistent way to interrupt an agent mid-task, get a human decision, and resume - especially when multiple people can be responsible for weighing in on the decisions. Vigilator provides a dashboard, where you and your teammates can join a virtual organisation - from here you can use our Py/TS SDK or simply make API calls yourself - then from here tool calls can be approved/reviewed/denied and agents can be monitored as they work. We provide a workload management system meaning that you can have multiple team members at once handling interrupts and watching agents as they work. Vigilator is still early, while Vigilator is framework agnostic I've primarily focused on adding support and examples for Mastra, Langchain Deep Agents and Langgraph - and I would love feedback, particularly from people running agents in production at scale. try it: https://vigilator.ai Docs: https://docs.vigilator.ai (conceptual overview: https://docs.vigilator.ai/concepts) If you would like to get in contact feel free to shoot me an email at [email protected]