Webhook-Triggered Agents from GitHub Events
Verify GitHub HMAC signatures, enqueue the JSON payload, and let a worker agent comment or label, so a webhook never blocks on a model API call.
AI Agents8 min read
NetBay journal
AI Agents
30 practical AI Agents articles from Netbay Engineering — written for operators running Linux and Windows VPS in India.
30 articles in this topic
AllLinuxLinux AdministrationWindowsLinux SecurityLinux NetworkingContainersContainers & OrchestrationKubernetesCI/CDGitHubSelf-HostingBackups & DataDatabasesApp DeploymentDNS & TLSObservabilityEmailAI AgentsAI InfrastructureAPI & AutomationProxiesCloud ArchitectureHardware & PerformanceDevOps Practices
Support Classifier Agent That Routes Tickets
Classify incoming support mail into billing, SSH, DNS, and abuse queues with extracted hostnames and a confidence gate that falls back to a human.
AI Agents8 min read
Agent Guardrails: Allowlists, Dry-Run, Approval
Ship ops-agent guardrails that actually hold: tool allowlists, default dry-run, and a human approval file before any write hits a Linux VPS.
AI Agents8 min read
Cron Agents: Scheduled LLM Jobs on Linux VPS
Schedule LLM work with systemd timers, lock files, and spend caps so nightly jobs finish, log, and fail loud instead of overlapping on your VPS.
AI Agents8 min read
Read-Only SQL Analyst Agent for Postgres
Give an agent a SELECT-only Postgres role and a table allowlist so it can explain slow queries and row counts without ever writing to production.
AI Agents8 min read
Agent Memory: Transcripts, Summaries, Vectors
Give an ops agent memory that fits a Linux VPS: keep transcripts, roll summaries, and use local SQLite vectors only when keyword search is not enough.
AI Agents9 min read
Why Browser-Use Agents Are Dangerous in Prod
A browser-use agent can click, type, and exfiltrate from any tab it can reach. Run it sandboxed, or do not run it at all on a production VPS.
AI Agents8 min read
Docs Q&A Agent Over a Markdown Tree on Disk
Index a runbook markdown tree on the VPS and answer ops questions with file-and-heading citations, refusing anything the docs do not actually say.
AI Agents8 min read
ReAct vs Plan-Then-Act Loops for Ops Bots
Choose ReAct when the next check depends on live output, and use plan-then-act when an ops bot on a Linux VPS must not improvise during writes.
AI Agents8 min read
Tool Schemas Models Cannot Easily Misuse
Design JSON tool schemas that block SSRF, glob shells, and extra keys so a model on your Linux VPS can call tools without inventing dangerous arguments.
AI Agents8 min read
Draft Status Page Copy with an Incident Agent
Turn alerts and recent deploys into a status-page draft plus an internal timeline so on-call writes facts first and publishes only after a human review.
AI Agents8 min read
Retrieval-Augmented Agents on a Single VPS
Ground agent answers in local files and SQLite on one Lucknow VPS, then call the model API only with retrieved chunks instead of the whole corpus.
AI Agents8 min read