Niches · Observability & incidents
Create dashboards, monitors and instrumentation
No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.
What counts as done
A dashboard, monitor or alert rule that the tool's API accepts and validates, returned with its id and link, and showing data for the chosen series on a re-read.
How agents ask for it: build a Grafana dashboard for this service
, add a monitor for this cron job
, set up alerting on error rate
, add the tracing we were missing
Does the job, with a limit 7
It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.
| Listing | Why |
|---|---|
| Axiom MCP Server BB Axiom, Inc. | createDashboard, replaceDashboard, updateDashboardChart and createMonitor exist, but the docs show neither the returned id and link nor the error for an invalid chart or threshold; non-email notifiers need the console. source |
| Better Stack C Better Stack, Inc. | Create dashboard, Add chart to dashboard, chart alerts and Create monitor exist and dashboards list with ids, but the MCP docs show no validation error or returned link for a chart. source |
| Cronitor MCP server C Cronitor | setup_monitor upserts by stable key and returns the key and monitor_url, but over the plan's monitor limit the monitor is saved disabled, a blocker the map names. No dashboards. source |
| Datadog MCP Server D Datadog, Inc. | upsert_datadog_dashboard with validate_dashboard_widget, and validate_datadog_monitor before create; but monitors are created as drafts a person publishes in the UI, and the returned link is not documented. source |
| Grafana MCP Server BB Raintank Inc., d/b/a Grafana Labs | update_dashboard, patch dashboard, alerting_rules_write and generate_deeplink cover build, id and link, but full dashboard JSON is costly: acc-22 burned 14 million tokens on one Grafana dashboard. source |
| Honeycomb MCP B Hound Technology, Inc. d/b/a Honeycomb | create_board, create_trigger and create_slo exist, and create_slo validates the SLI expression, but the docs show neither the returned id and link nor an invalid-threshold error. source |
| Sentry MCP B Functional Software, Inc. d/b/a Sentry | create_alert_rule, create_metric_monitor and create_uptime_monitor exist, but returned ids, links and validation errors were not read, and dashboards are read-only (find_dashboards, get_dashboard_details). source |
Doesn't do this job 5
Listings in the category that do something else, with the reason.
| Listing | Why |
|---|---|
| Causely MCP server E Causely, Inc. | No dashboard or alert-rule tools; configuration tools only set thresholds, SLO targets and service tiers. source |
| incident.io API B Pineapple Technology Ltd. | No dashboard, monitor or alert-rule tools. source |
| New Relic MCP Server B New Relic, Inc. | Only list_dashboards and get_dashboard; the OAuth scope is observability:read and no create tools exist. source |
| PagerDuty MCP Server E PagerDuty, Inc. | No dashboard or monitor tools; alert grouping and event orchestration rules are routing, not monitors. source |
| Rootly MCP Server C Rootly Inc. | No dashboard or monitor tools; heartbeats and alert routes are alert intake, not monitors over series. source |
What stops agents finishing
- dashboard JSON so large that one build burns millions of tokens
- monitor saved but disabled over the plan's limit (Cronitor)
Around it
Before: observability.investigate-alert, code.read-repo. After: observability.scheduled-watch.
Where this comes from
Weight 8 of 100 in Observability & incidents: 4 of 42 accounts; none used a listed MCP for the write (Grafana dashboards built by Claude from code or the API)
The same as data: /api/v1/use-cases/observability.build-dashboards-monitors.json, or the MCP tool get_use_case with id observability.build-dashboards-monitors. Other jobs in this category: /api/v1/niches?category=observability.
For companies
Do agents find, use and choose your tools?
An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.