OpenTelemetry Skill
Skill configures OpenTelemetry: collectors, sampling, security and telemetry from AI agents, without flooding the context.
- 55 ★ on the repository
- Last commit
Minimal access
Why this level
- The skill provides reference knowledge, but generated collector configs must be checked before rollout.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/opentelemetry-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill provides reference knowledge, but generated collector configs must be checked before rollout.
This is third-party code. Review the repository files before installing.
What it does
OpenTelemetry Skill teaches an agent to think about observability before it writes configuration. `SKILL.md` acts as a router: it forces an assessment of throughput, cardinality and resiliency first and loads detailed references only by topic. They cover architecture, the collector, instrumentation, security, sampling, the OTTL transformation language and compatibility. There are deployment guides for Kubernetes, ECS, Docker and virtual machines, guards against metric explosion, PII redaction and TLS setup. A separate reference covers monitoring AI agents themselves: Claude Code, Gemini CLI, Codex, GitHub Copilot and others. The skill is tested on more than 20 scenarios.
Who it is for. SRE engineers and developers deploying the OpenTelemetry Collector who want an agent to produce working configs instead of invented ones.
Good fit when
- A tail sampling gateway is deployed in Kubernetes with `traceID` load balancing.
- Metric cardinality is growing and cost must be capped.
- Personal data is stripped from logs and traces before they reach storage.
- Claude Code or Codex telemetry is sent into a shared collector.
Not a fit when
- You need a ready monitoring product rather than OpenTelemetry knowledge.
- Your stack does not use OpenTelemetry.
- You plan to roll out the configuration without testing it in a staging environment.
Example request
I need an OpenTelemetry gateway in Kubernetes for tail sampling with error and latency policies. Ask about throughput, size the replicas and give validation commands.Limitations
Installing through `npx skills add` needs Node.js 22.20 or newer and npm. The `gen_ai.*` namespace in OpenTelemetry is still experimental and attribute names may change. Trace coverage varies among AI agents: Claude Code emits logs and metrics while traces are still beta, so check your version. Tail sampling memory scales with in-flight traces and OTTL regex transforms affect p99 latency. Test configurations under realistic load.
How to disable. Delete the skill folder from `.claude/skills/` or `.agents/skills/`. Generated collector configs stay in your repository and are removed separately.
FAQ
How do I check without installing?
Run `npx skills add o11y-dev/opentelemetry-skill --list`: the installer shows the found skill and installs nothing.
What is progressive loading?
`SKILL.md` stays a short instruction while detailed references load only when a topic triggers them. That keeps the whole documentation out of the context.
Which deployment platforms are covered?
Kubernetes with a DaemonSet, gateway or sidecar choice, ECS on EC2 and Fargate, Docker Compose and plain virtual machines.
Does it cover AI agents?
Yes. `references/ai-agents.md` describes how to collect telemetry from Claude Code, Gemini CLI, Codex, GitHub Copilot and others, with a caveat about partial trace coverage.
Which scenarios are tested?
The README states a TDD-based methodology and more than 20 evaluation scenarios. Running them is described in a separate testing section.
Discussion
No questions or reviews yet.
Related
SkillsOpenTelemetry Agent Skills helps agents instrument applications, configure Collector and check versions against primary sources.
Four OpenTelemetry skills covering application instrumentation, semantic conventions, the Collector and the OTTL language.
KubeShark skill: the agent finds the failure mode first, then writes manifests, Helm and Kustomize to official practice with a rollback plan
The official Grafana skill set: PromQL, Loki, Tempo, Alloy, k6, dashboards and alerting for Claude Code, Cursor and Codex
A large base of DevOps, security, infrastructure and compliance skills for AI agents
31 DevOps skills for Claude Code and Codex: config generators, validators and a Kubernetes debugger for Terraform, Helm, CI and Docker
README in short
The README describes the router principle, skill structure, references, platform guides, example requests, testing and known limitations. The repository is published as a Tessl tile and in the skills.sh directory.