Tools Reference

All tools accept an optional project_id parameter to override the default project scope.

Logs

`query_logs`

Execute a ClickHouse SQL query against otel_logs.

Parameter	Type	Required	Description
`query`	string	Yes	ClickHouse SQL query
`time_from`	string	Yes	Start time (RFC3339)
`time_to`	string	Yes	End time (RFC3339)
`limit`	integer	No	Max rows (default 100)

`search_logs`

Full-text search across log message bodies.

Parameter	Type	Required	Description
`text`	string	Yes	Search query
`severity`	string[]	No	Filter by severity (ERROR, WARN, INFO)
`service`	string	No	Service name filter
`time_from`	string	Yes	Start time (RFC3339)
`time_to`	string	Yes	End time (RFC3339)
`limit`	integer	No	Max results (default 50)

`list_log_services`

List all services that emitted logs in a time range.

`log_stats`

Aggregated log statistics grouped by severity, service, or other fields.

`log_patterns`

Discover recurring log patterns using automatic clustering.

Traces

`query_traces`

Execute a ClickHouse SQL query against otel_traces.

Parameter	Type	Required	Description
`query`	string	Yes	ClickHouse SQL query
`time_from`	string	Yes	Start time (RFC3339)
`time_to`	string	Yes	End time (RFC3339)
`limit`	integer	No	Max rows (default 100)

`get_trace`

Get the full trace tree for a specific trace ID.

Parameter	Type	Required	Description
`trace_id`	string	Yes	The trace ID to retrieve

`search_traces`

Search for traces by service, operation, latency, and status.

Parameter	Type	Required	Description
`service`	string	No	Service name filter
`operation`	string	No	Operation/span name filter
`min_latency`	string	No	Minimum duration (e.g. `500ms`, `1s`)
`max_latency`	string	No	Maximum duration
`status`	string	No	Span status: `OK`, `ERROR`, `UNSET`
`time_from`	string	Yes	Start time (RFC3339)
`time_to`	string	Yes	End time (RFC3339)
`limit`	integer	No	Max traces (default 50)

`list_trace_services`

List services that emitted trace spans.

`list_trace_operations`

List operations for a given service.

`trace_stats`

Aggregated trace statistics (p50/p95/p99 latency, error rates, throughput).

Metrics

`query_metrics`

Execute a ClickHouse SQL query against otel_metrics.

`list_metric_names`

List available metric names with optional prefix and service filters.

`get_metric_metadata`

Get metadata for a metric (type, description, unit).

`list_metric_label_names`

List label names for a metric.

`list_metric_label_values`

List values for a specific label of a metric.

`metric_stats`

Time-series statistics with configurable aggregation step.

Dashboards

`list_dashboards`

List dashboards with optional search query.

`get_dashboard`

Get full dashboard details including panel configurations.

`get_dashboard_panels`

Get the panel configurations and queries for a dashboard.

`create_dashboard` (write)

Create a new dashboard.

`update_dashboard` (write)

Update a dashboard's name or description.

`delete_dashboard` (write)

Delete a dashboard by its ID.

Incidents

`list_incidents`

List incidents with filters for status, severity, service, and time range.

Parameter	Type	Required	Description
`status`	string	No	`open`, `acknowledged`, `resolved`, `closed`
`severity`	string	No	`critical`, `high`, `medium`, `low`
`service`	string	No	Affected service name
`time_from`	string	No	Start time (RFC3339)
`time_to`	string	No	End time (RFC3339)
`limit`	integer	No	Max incidents (default 50)

`get_incident`

Get full details of a specific incident.

`get_incident_timeline`

Get the activity timeline for an incident.

`create_incident` (write)

Create a new incident with title, severity, and affected service.

`update_incident` (write)

Update incident status or severity.

`add_incident_activity` (write)

Add a note, status change, or action to an incident's timeline.

Alerts

`list_alert_rules`

List alert rules with status and severity filters.

`get_alert_rule`

Get full alert rule details including query and conditions.

`list_alert_history`

View alert firing and resolution history.

`list_contact_points`

List notification contact points (email, Slack, PagerDuty, etc.).

`create_alert_rule` (write)

Create a new alert rule.

`update_alert_rule` (write)

Update an alert rule.

`delete_alert_rule` (write)

Delete an alert rule.

AI

`chat_to_query`

Convert natural language into a ClickHouse query against Obtrace telemetry data.

Parameter	Type	Required	Description
`question`	string	Yes	Natural language question
`query_type`	string	No	Target: `logs`, `traces`, `metrics`

`root_cause_analysis`

AI-powered root cause analysis correlating logs, traces, and metrics.

Parameter	Type	Required	Description
`incident_id`	string	No	Incident to analyze
`trace_id`	string	No	Trace to analyze
`service`	string	No	Affected service
`time_from`	string	Yes	Start of issue window (RFC3339)
`time_to`	string	Yes	End of issue window (RFC3339)
`symptoms`	string	No	Description of observed symptoms

`summarize`

Generate an AI summary of system health for a time window.

`autofix` (write)

Generate code fix suggestions for errors found in telemetry data.

Projects

`list_projects`

List all accessible projects.

`get_project`

Get project details including apps and environments.

`list_apps`

List applications within a project.

`list_teams`

List organization teams.

`list_users`

List organization users.

Replay

`list_replay_sessions`

List session replay recordings with filters.

`get_replay_session`

Get session metadata and summary.

`get_replay_events`

Get the event timeline for a session (clicks, navigations, network, errors).

`get_replay_errors`

Get all errors from a replay session.

Search

`global_search`

Search across all Obtrace data types with a single query.

`generate_deeplink`

Generate a deep link URL to a specific view in the Obtrace UI.

Parameter	Type	Required	Description
`type`	string	Yes	`logs`, `traces`, `metrics`, `dashboard`, `incident`, `replay`
`resource_id`	string	No	Resource ID
`query`	string	No	Pre-filled query
`time_from`	string	No	Start time
`time_to`	string	No	End time

Tools Reference

On this page