Waspen
AI security monitoring that runs entirely on your machine.
Local LLMs. No telemetry. No cloud required. Free for up to 50 monitoring sources.
Want launch notes and build announcements?
Share your email if you want occasional Waspen updates: beta build availability, release notes, and major feature announcements. No telemetry, no sales drip.
How It Works
A continuous four-step loop. No babysitting required.
Monitor
Waspers are monitoring sources. Each one polls an endpoint, metric, log, Prometheus query, syslog stream, or dashboard on your schedule.
Detect
The rule engine evaluates thresholds, trends, and LLM-driven anomaly detection in a single pass. Not just numbers — patterns.
Decide
Multi-step pipelines with optional approval gates. Auto-remediate what you trust. Escalate what you don't.
Act
Fire webhooks, push ChatOps notifications to Slack, Teams, or Discord, or trigger automated remediation scripts.
Built for the Way You Actually Work
Everything you need to monitor infrastructure without stitching together six different tools.
Rule Engine
Three detection modes in one: expression-based thresholds for known limits, trend analysis for gradual drift, and LLM-driven anomaly detection for the weird stuff no formula catches.
- Expression rules with cooldowns
- Trend detection over time windows
- LLM anomaly detection for unstructured data
Eyes — Visual Monitoring Pro
Headless browser screenshots of dashboards and status pages. DOM text grounding for accuracy. AI analysis with evidence bundles — screenshot, page text, Wasper metadata, and model commentary all together.
- True screenshot monitoring via Chromium
- DOM text grounding to reduce hallucinations
- Latest-state cache, image retention controls
Ask Wasper — AI Chat
Conversational AI with real-time awareness of your system state, incidents, and Wasper data. Ask what's happening right now, get plain-language summaries, or request a quick incident explanation.
- Context-aware: knows your live status
- Runs against your local Ollama endpoint
- Incident Explain gives plain-language breakdowns (Pro)
Builder — Workflow Canvas Pro
Drag-and-drop if-this-then-that workflow builder with a visual canvas, zoom/pan, dry-run preview, and risky-action approval gates. Build complex automation without writing YAML.
- Visual workflow canvas with inspector
- Dry run: preview before saving
- Approval gates for risky actions
Tiered Starter Job Catalog
One click to enable real monitoring — no configuration marathon before you see value. Deploy, toggle on the jobs you care about, and start getting alerts in minutes.
- Disk usage, CPU load, memory
- Open ports, process health
- SSL certificate expiry
- Log file pattern detection
Offline Licensing — Ed25519
No license server. No phone-home. No "please connect to the internet to validate your purchase." Waspen uses Ed25519 cryptographic signatures that verify locally. Deploy in an air-gapped lab, a closet server, anywhere.
- 100% offline verification
- Ed25519 signed license files
- Free tier needs no license at all
Seven Starter Jobs. Infinite Extensibility.
Waspen ships with seven monitoring jobs ready to go. No YAML marathon before you see value — deploy, start monitoring, customize later. Pro unlocks a full security pack, homelab pack, and integrations for Plex, Proxmox, AdGuard Home Assistant, Ollama, and more.
- Disk Usage — catch drives filling up before they crash your services
- CPU Load — spot runaway processes and sustained spikes
- Memory — track consumption and detect leaks early
- Open Ports — know what's listening and what shouldn't be
- Process Health — verify critical services are running
- SSL Certificates — never get caught by an expired cert again
- Log Files — LLM-powered pattern analysis and anomaly detection
Local AI. Zero Egress.
Waspen is built around local-first AI. Configure an Ollama endpoint once and every analysis, anomaly detection, and advisory board runs on your hardware. Nothing leaves your machine by default.
For Pro users who need heavier models, Waspen's External AI Egress Guard redacts secrets and PII before any cloud API call, with a metadata-only privacy receipt so you always know what was sent.
- Ollama supported out of the box
- Task-specific small models (phi3, qwen2.5, granite3.1)
- AI work scheduler prevents model contention
- External AI + Egress Guard available in Pro
Simple, Honest Pricing
Start free. Upgrade when you need the powerful stuff.
Part of the STING Ecosystem.
Waspen is part of the AlphaBytez bee ecosystem. Waggles bridges your Waspen alerts directly into Slack, Teams, and Discord through STING ChatOps — the same notification layer used by Hive enterprise teams. Same hive, different job.
- Waggles bridge for Slack, Teams, Discord alerts
- Platform-native message formatting
- Shared STING notification infrastructure
- Natural upgrade path to Hive for enterprise teams
The guard that never sleeps.
50 free Waspers. Local-first monitoring. Zero telemetry. Point it at your Ollama endpoint and let it watch.
Common Questions
Does Waspen require an internet connection?
No. Waspen runs entirely offline. The binary is self-contained with embedded SQLite, and licensing validates locally with Ed25519 signatures. For AI analysis, point it at a local Ollama endpoint — your data never leaves. The only time you'd need internet is if you opt into a remote LLM provider like OpenAI, and that's always your explicit choice.
What do I need to run the AI features?
An Ollama installation on the same machine or your local network. Waspen recommends small, fast models like phi3:mini, qwen2.5:3b, or granite3.1-moe:3b — they're effective for the monitoring tasks Waspen sends them, and they run comfortably on a modern laptop or homelab machine.
What's the difference between Free and Pro?
Free gives you up to 50 Waspers (3 rules each), 14 days of incident history, starter job templates, local AI chat, native notifications, and single-action webhook responses. That's genuinely useful for personal and small-scale monitoring.
Pro is planned to unlock Eyes (visual dashboard monitoring via headless browser), Advisor AI boards (curated insight panels), the Builder drag-and-drop automation canvas, the full security and homelab job packs, Pro integrations (Proxmox, Plex, AdGuard, etc.), external AI with Egress Guard, full pipeline execution with approval gates, unlimited Waspers — and email support. Final purchase options are still being validated. See the full comparison →
What's a Wasper?
A Wasper is a configured monitoring source — one HTTP endpoint check, one set of system metrics, one log file, or one visual dashboard source. Waspers collect signals on a schedule and write snapshots. Rules attach to a Wasper and decide when those snapshots become incidents or actions. The Free tier includes 50 Waspers, which is enough for a well-monitored homelab or personal server setup. Pro removes the limit entirely.
How will Pro licensing work?
The current plan supports signed license.json files that Waspen can verify locally with no network call. Direct purchase, subscription, and app-store managed paths are still being validated before public pricing is finalized.
Is Waspen open source?
No. Waspen is a proprietary product by AlphaBytez LLC. It is not open source. There is no public repository. Contact us with licensing or access questions.