whatsmyagentdoing vs agent-observability
Local coding-agent visibility compared
whatsmyagentdoing shows local coding-agent sessions in a live office-style dashboard, while agent-observability collects telemetry and presents it in Grafana. The first focuses on glanceable status across many agent tools; the second focuses on metrics, logs, traces, and self-hosted observability for Claude Code and Codex.

whatsmyagentdoing: See Coding Agents in a Live Office
A Claude Code and Codex skill that shows what local coding agents are doing in a live office, with an optional Three.js 3D view. It also reads sessions from several other agent tools and runs locally with Node.js.

agent-observability: Monitor AI Coding Agents Locally
A self-hosted OpenTelemetry stack for monitoring Claude Code and OpenAI Codex. It routes telemetry to Prometheus, Loki, and Tempo, then presents usage, performance, and activity in Grafana dashboards.
| whatsmyagentdoing | agent-observability | |
|---|---|---|
| Language | JavaScript | Go Template |
| License | MIT | — |
| Stars | 375 | 14 |
| Forks | 51 | 1 |
| Last analyzed | Nov 4, 2026 | Oct 4, 2026 |
Key differences
- whatsmyagentdoing supports a range of agent tools, while agent-observability covers Claude Code and OpenAI Codex.
- whatsmyagentdoing presents session status in 2D, optional 3D, and list views; agent-observability routes telemetry to Prometheus, Loki, and Tempo for Grafana dashboards.
- whatsmyagentdoing runs locally with Node.js and reads session stores; agent-observability requires operating an OpenTelemetry Collector and supporting services.
- agent-observability offers Docker Compose, Kubernetes with Kustomize, and Helm deployment paths; whatsmyagentdoing is installed and used locally as a skill.
- whatsmyagentdoing is written in JavaScript and has an MIT license; agent-observability uses Go Template.
- The supplied metadata lists 375 stars and 51 forks for whatsmyagentdoing, compared with 14 stars and 1 fork for agent-observability. Their maintenance needs also differ: whatsmyagentdoing has contributions concentrated mostly in one person, while agent-observability requires maintaining its observability stack.
Choose whatsmyagentdoing if you…
- want a live, glanceable view of several local agents, including questions, errors, and work status.
- need support for agent tools beyond Claude Code and Codex, with optional 2D or 3D office views.
- prefer a local Node.js dashboard with session details and optional privacy mode.
Choose agent-observability if you…
- need metrics, logs, and traces for Claude Code and Codex in Grafana dashboards.
- want to investigate usage, tool activity, latency, and session events with Prometheus, Loki, and Tempo.
- can operate the supporting services yourself or deploy them alongside existing Kubernetes services.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.