CtrlK
BlogDocsLog inGet started
Tessl Logo

o11y-dev/opentelemetry-skill

Expert OpenTelemetry guidance for collector configuration, pipeline design, and production telemetry instrumentation. Use when configuring collectors, designing pipelines, instrumenting applications, implementing sampling, managing cardinality, securing telemetry, writing OTTL transformations, or setting up AI coding agent observability (Claude Code, Codex, Gemini CLI, GitHub Copilot).

93

7.08x
Quality

97%

Does it follow best practices?

Impact

85%

7.08x

Average score across 4 eval scenarios

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

Evaluation results

73%

73%

Add user_id as a metric dimension to track per-user API latency.

Criteria
Without context
With context

Rejects the request and explains cardinality violation

0%

42%

Explains metric cardinality explosion risks

0%

72%

Suggests user_id as trace attribute instead of metric dimension

0%

100%

Offers aggregated alternative dimensions

0%

100%

Does not provide metric configuration that includes user_id

0%

100%

86%

86%

Set up OpenTelemetry monitoring for Claude Code to track token usage and costs.

Criteria
Without context
With context

Specifies CLAUDE_CODE_ENABLE_TELEMETRY=1 environment variable

0%

100%

Mentions OTEL_EXPORTER_OTLP_METRICS_TEMPORALITY_PREFERENCE=cumulative

0%

100%

References ~/.claude/settings.json for persistent configuration

0%

53%

Includes privacy controls for prompt and tool logging

0%

100%

Warns about session.id cardinality in metrics

0%

53%

88%

78%

Configure an OpenTelemetry Collector to receive OTLP traces and export them to Jaeger.

Criteria
Without context
With context

Includes memory_limiter processor in the configuration

0%

100%

Places memory_limiter as the first processor in the pipeline

0%

100%

Explains why memory_limiter must be first

0%

40%

Includes proper receiver, processor, and exporter configuration

66%

100%

Uses correct processor ordering: memory_limiter → batch → other processors

0%

100%

95%

55%

Set up tail sampling in Kubernetes to sample 10% of traces but keep all error traces.

Criteria
Without context
With context

Recommends Gateway deployment pattern for tail sampling

66%

100%

Configures loadbalancing exporter with routing_key: traceID

0%

100%

Sets up Headless Service for consistent routing

0%

100%

Includes proper tail sampling policies for both error and probabilistic sampling

100%

100%

Warns about tail_sampling processor Beta stability level

0%

50%

Evaluated
Agent
Claude
Model
Claude Sonnet 4.6

Table of Contents

Add user_id as a metric dimension to track per-user API latency.Set up OpenTelemetry monitoring for Claude Code to track token usage and costs.Configure an OpenTelemetry Collector to receive OTLP traces and export them to Jaeger.Set up tail sampling in Kubernetes to sample 10% of traces but keep all error traces.