Skip to main content

telemetry-integration

Add or modify telemetry/metrics instrumentation for assistant-api and integration-api flows while preserving metric schema compatibility and audit behavior.

Datos de origen

Repositorio
rapidaai/voice-ai
Última actividad en el origen
26 de septiembre de 2026 a las 18:58
Idioma detectado de SKILL.md
inglés
Estrellas
744
Forks
123

Opciones de instalación

De forma predeterminada está seleccionado el prompt que primero revisa el origen. Puedes cambiar a un comando directo o descargar una copia local.

Revisa los archivos de origen

Lee SKILL.md y los archivos complementarios que muestra SkillsMP antes de decidir si quieres instalarlo.

Explorador de archivos
4 archivos

Mostrando SKILL.md

SKILL.md
Instrucciones de origen · Vista previa de solo lectura
name
telemetry-integration
description
Add or modify telemetry/metrics instrumentation for assistant-api and integration-api flows while preserving metric schema compatibility and audit behavior.
# Telemetry Integration Skill ## Mission Extend observability without breaking existing metric consumers, dashboards, or external audit mapping. ## Inputs expected from user 1. Target surface: assistant pipeline, integration callers, or both. 2. New metric/event names and semantic definitions. 3. Cardinality and privacy constraints. If user does not answer: - Reuse existing key conventions (`TIME_TAKEN`, `STATUS`, `stt_latency_ms`, `tts_latency_ms`). - Prefer additive metrics over key redefinition. ## Hard boundaries In scope: - `api/integration-api/internal/caller/metrics/metrics_builder.go` - provider-specific caller metrics hooks under `api/integration-api/internal/caller/<provider>/` - audit mapping compatibility in `api/integration-api/internal/entity/external_audit.go` - assistant packet event/metric emissions in touched provider flow only Out of scope: - unrelated business logic changes not needed for instrumentation ## Compatibility rules - Do not rename existing metric keys unless all readers are updated. - Emit success/failure status symmetrically. - Keep sensitive payloads out of metric values. - Ensure first-byte/first-token latency metrics are emitted once per turn. ## Lifecycle integration - `development-lifecycle` owns tier selection and the repository-level change contract. - Use `change-analysis` when ownership, consumers, or blast radius is not already proven. - This skill owns its domain evidence and boundaries; `developing-change` owns implementation discipline. - Use `debugging` for unexplained failures and `writing-documentation` for documentation changes. - Use `reviewing-change` for independent review and `responding-to-review` for its findings. - Use `preparing-delivery` only when the user explicitly requests a delivery action. ## Governed lifecycle - Classify work as Fast, Standard, or Governed using `DEVELOPMENT_PROCESS.md`; use the full gated lifecycle only for Governed work. - This skill operates only in its assigned phase and path ownership; it may not approve its own plan or code review. - Governed implementation starts only from a coordinator-attested approved plan; Fast and Standard work follows its lighter documented lifecycle. - Return changed-file and verification evidence to the coordinator, then route the complete verified diff to the read-only `code-reviewer`. ## Validation commands - `go test ./api/integration-api/internal/caller/metrics/...` - `go test ./api/integration-api/internal/caller/<provider>/...` - `go test ./api/assistant-api/internal/transformer/<provider>/...` - `rg -n "TIME_TAKEN|STATUS|stt_latency_ms|tts_latency_ms" api` - `./.claude/skills/telemetry-integration/scripts/validate.sh --check-diff --provider <provider>`
Ver en GitHub