[Telemetry] Build privacy-reviewed Hevy usage and reliability dashboards

[chrisdoc]

Summary

Create a small set of Sentry/Honeycomb dashboards that answer product and reliability questions using only the approved bounded telemetry fields.

Questions to answer

Product usage

• Which Hevy feature areas are used most?
• What is the read/write split?
• Which tools are used as discovery or search entry points?
• How often are workflows such as training summaries and routine discovery used?
• How frequently do users create or update workouts, routines, folders, and exercise templates?

User experience

• Which tools have the highest returned-error and thrown-error rates?
• Which endpoints/statuses cause the most friction?
• Are 404s concentrated in stale IDs or pagination?
• Are 409s concentrated in duplicate/conflicting writes?
• How often do retries, rate limits, or timeouts occur?
• Which MCP client versions correlate with failures?

Performance

• Tool duration by feature and operation
• Hevy API duration by normalized endpoint and method
• Workflow pages scanned and cache hit/miss rate
• Result-size buckets and their relationship to latency

Required privacy review

Before publishing dashboards:

• Confirm prompt/input/output capture remains disabled.
• Confirm no raw queries, IDs, titles, notes, descriptions, dates, or measurements are used.
• Confirm all metric dimensions have an explicit bounded value set.
• Confirm user-level views are limited to short-lived troubleshooting and aggregate unique-user counts.
• Record the approved retention period and dashboard access owners.

Acceptance criteria

• A documented dashboard set exists for product usage, tool reliability, API reliability, and performance.
• Every panel lists its source span/metric and dimensions.
• Panels use only approved privacy-safe fields.
• A telemetry data dictionary is checked into the repository or linked from the issue.
• A review checklist or automated test prevents reintroducing raw input/output capture.

[coderabbitai]

🔗 Related PRs

#443 - feat: instrument with OpenTelemetry for Sentry + Honeycomb dual export [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!