production-debugging

yc-360 is a lightweight diagnostic script that captures comprehensive troubleshooting artifacts, including application logs, GC logs, thread dumps, heap dumps, and OS-level metrics across a wide range of environments, including Linux, Windows, macOS, Docker, Kubernetes, OpenShift, and AWS Fargate.

🐺 Time-travel debugger for distributed systems. Record production. Replay locally. Byte-for-byte identical.

Temporal debugging skill - time-travel through git history to debug bugs in their original context

Read-only MCP diagnostic server for Spring Boot — point any AI agent at your running service and ask it to debug production issues.

A zero-allocation, high-performance custom Go slog.Handler for package-scoped logging. Dynamically adjust log levels per package in production at runtime without application restarts.

Agentry MCP and Claude plugin/skill for agent-native analytics, error logging, and deploy attribution.

htop for Go goroutines—a live terminal visualizer and leak detector that groups pprof stacks, tracks growth, and pinpoints stuck goroutines in local or remote processes.

Turn production bugs into reproducible local executions. Capture the execution, not the environment — replay it locally, diff production vs local, and verify the fix. Node.js · PostgreSQL · HTTP. Currently in a Phase 0 technical spike.

Evidence-first AI Production Support Engineer that explains why software failed or got slow using traces, logs, metrics and deterministic RCA.

Package failures into redacted evidence bundles and compare what changed across local, CI, and OpenTelemetry incidents.

Agent-first control plane for PostHog — source-backed analytics, production investigation, and bounded self-healing and self-improving software workflows.

Agentry skill for agent-native analytics, error logging, and deploy attribution.

GitHub profile: paid AI-agent ops mini-sprints for bug triage, PR risk review, deploy smoke, and automation.

Paid AI-agent operations mini-sprints: production bug triage, PR risk review, AI-code cleanup, deploy smoke, automation, plus Korean templates.

C#/.NET debugging case study demonstrating diagnosis and remediation of a silent data-integrity defect involving async execution and persistence.

Interactive simulator showing how AI agents fail silently — HTTP 200, valid JSON, zero error signals, real task never executed.

Rookout — independent third-party profile of a public API surface, by API Evangelist. Rookout is a live application debugging platform founded in 2017 in Tel Aviv, Israel that lets developers set non-breaking breakpoints in running production, staging, and development applications to capture full-stack debug data — variables, stack traces, logs, an

Hyperprobe — independent third-party profile of a public API surface, by API Evangelist. Hyperprobe is a runtime data platform for AI-native software development, built by a San Francisco team in Y Combinator's Summer 2026 batch. It lets developers and coding agents (Cursor, Claude Code, Codex, opencode, Windsurf) set live, non-breaking breakpoints