Analyzes Copilot and agentic workflow OpenTelemetry token and execution metrics from JSONL exports.
npx skills add https://github.com/dotnet/maui --skill otel-queries
Please help me install this Agent Skill. Skill: otel-queries Source: https://github.com/dotnet/maui/tree/main/.github/skills/otel-queries Command: npx skills add https://github.com/dotnet/maui --skill otel-queries
Investigating Copilot and agentic workflow telemetry is difficult when token usage, model usage, and execution timing data are scattered across OpenTelemetry exports. This Skill provides guidance for querying and interpreting that telemetry so you can understand where tokens and time are spent.
COPILOT_OTEL_FILE_EXPORTER_PATH JSONL files when an OTEL export is available.Analyze the Copilot OTEL export file and summarize token usage by model and agent step.
High-intent search queries and answers about installing and using this skill.
Inspect the exported OTEL JSONL file locally and summarize token counts grouped by model, agent step, and request type. This reveals which steps and models consume the most tokens in an agentic workflow.
Read the JSONL export line by line, where each line is a separate telemetry record. Parse each record to extract token counts, model identifiers, request types, and execution timing fields for aggregation.
Treat missing or malformed OTEL records as diagnostic data rather than workflow failures. Gaps in telemetry can themselves indicate instrumentation or export issues worth investigating.
No, workflow-only telemetry analysis changes do not require Android or product validation. The analysis operates purely on exported telemetry data, independent of platform-specific builds.
Examine the execution timing fields in the OTEL export and compare durations across agent steps and request types. Outliers in timing highlight steps that dominate overall workflow latency.
dotlab-hqInstrument applications and query LangSmith traces to debug LLM workflows.
patricio0312revOptimize LLM applications for cost and latency via caching, model selection, batching, and prompt optimization.
Cogni-AI-OUAudit GitHub Agentic Workflow runs and logs to diagnose failures.
BagelHoleInstrument applications and infrastructure with OpenTelemetry for unified traces, metrics, and logs.