feat(dashboards): ship Grafana dashboard JSON for agent obs signals (RIG-2826) - #670
Open
rigel-mintaka wants to merge 2 commits into
Open
feat(dashboards): ship Grafana dashboard JSON for agent obs signals (RIG-2826)#670rigel-mintaka wants to merge 2 commits into
rigel-mintaka wants to merge 2 commits into
Conversation
…RIG-2826) Task T5 of the frozen observability architecture record (`docs/designs/platform/compass-observability-architecture/design.md`, #656): ship prebuilt Grafana dashboards in-repo so an operator's Grafana — and Rigel's managed Grafana, "just like a user" — import them. Near-universal peer practice (GitLab/Temporal/Supabase/Coder all ship dashboard JSON). Two dashboards, bound only to signals the shipped `compass-agent` emits today: - `dashboards/compass-agent-transport.json` — 10 panels over the transport metrics (publish spine, frame sink, control source), Prometheus datasource. PromQL targets the OTel-Prometheus-mangled names of the verified constants in `packages/compass-agent/src/transport/otel-metrics.ts` (`.`→`_`, monotonic counters get `_total`; gauges keep the dotless name). - `dashboards/compass-agent-loop-traces.json` — 2 panels (invoke_agent root spans + execute_tool spans), Tempo datasource, TraceQL filtered to `service.name = compass-agent` (GenAI-semconv loop spans). Both declare their datasource as a templating variable (`${DS_PROMETHEUS}`/`${DS_TEMPO}`), never a hard-coded uid, so the JSON imports into any Grafana. No provisioning automation — import is the operator's one step (D3/T5). `dashboards/README.md` documents the rendered signals + import steps; server/runner panels are a noted follow-up (T4b). Spec-impact: none. Refs RIG-2826, #656 Co-authored-by: Matt Wilkinson <matt@rigel.build>
|
Compass engineering docs preview: https://compass-obs-rig-2826-dashboa.compass-eng-docs.pages.dev Deployed from Changed pages: |
Review low: the two frame-loss panels (`publish.trace_frames_lost`, `publish.priority_frames_lost`) used Grafana's built-in `fps` unit, which renders as `f/s` framerate and reads oddly for a dropped-frames-per-second rate. Switch both to `short` to match the sibling rate panels and avoid an axis-label misread. Query and data unchanged. Refs RIG-2826 Co-authored-by: Matt Wilkinson <matt@rigel.build>
rigel-mintaka
marked this pull request as ready for review
August 27, 2026 03:54
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Task T5 of the frozen observability architecture record (
docs/designs/platform/compass-observability-architecture/design.md, #656): ship prebuilt Grafana dashboards in-repo so an operator's Grafana — and Rigel's managed Grafana, "just like a user" — import them. Near-universal peer practice (GitLab/Temporal/Supabase/Coder all ship dashboard JSON).Two dashboards, bound only to signals the shipped
compass-agentemits today:dashboards/compass-agent-transport.json— 10 panels over the transport metrics (publish spine, frame sink, control source), Prometheus datasource. PromQL targets the OTel-Prometheus-mangled names of the verified constants inpackages/compass-agent/src/transport/otel-metrics.ts(.→_, monotonic counters get_total; gauges keep the dotless name).dashboards/compass-agent-loop-traces.json— 2 panels (invoke_agent root spans + execute_tool spans), Tempo datasource, TraceQL filtered toservice.name = compass-agent(GenAI-semconv loop spans).Both declare their datasource as a templating variable (
${DS_PROMETHEUS}/${DS_TEMPO}), never a hard-coded uid, so the JSON imports into any Grafana. No provisioning automation — import is the operator's one step (D3/T5).dashboards/README.mddocuments the rendered signals + import steps; server/runner panels are a noted follow-up (T4b).Spec-impact: none. Refs RIG-2826, #656
Co-authored-by: Matt Wilkinson matt@rigel.build