MCP server
Added 4 tools to the MCP server tool catalog:
- Get Infrastructure Resource Health History
- Get Infrastructure Resource Raw Data
- List Infrastructure Resource Types
- List Infrastructure Resources
See the changelog.
Added 14 tools to the MCP server tool catalog:
- Get case events
- Get Data Usage Query API capabilities
- Get saved view
- Get saved view folder
- List dashboard folders
- List dashboards
- List saved view folders
- List saved views
- Manage and inspect cases
- Create or delete dashboard folders
- Create, replace, validate, or delete dashboards
- Create, update, or delete saved view folders
- Create, update, or delete saved views
- Query billable data usage
See the changelog.
Added 3 tools to the MCP server tool catalog:
- Manage custom AI evaluation
- Manage AI evaluation
- Manage AI model pricing
Removed 13 tools:
- Add policy to application
- Create custom AI evaluation
- Create AI evaluation
- Delete AI evaluation
- Get AI application
- Get AI evaluation coverage
- Get AI evaluation
- List custom AI evaluations
- List custom AI evaluations for application
- Remove policy from application
- Set AI model pricing
- Update custom AI evaluation
- Update AI evaluation
See the changelog.
Coralogix available as a Claude Connector
The Coralogix MCP server is now available as a native connector in Claude. Find it under Claude → Customize → Connectors to query your logs, metrics, traces, RUM, alerts, and dashboards directly from Claude conversations — no additional setup beyond your existing MCP credentials. Learn more
MCP server v2 is now the default
Requests to the Coralogix MCP server without an mcp-version header now route to v2 automatically. If your integration depends on v1 behavior, send "mcp-version": "v1" in the request header to keep the previous routing. v1 remains available but is deprecated. Learn more
Coralogix in the Cursor Marketplace
The Coralogix Observability MCP Server is now available from the Cursor Marketplace. Install it under Cursor Settings → Plugins/Marketplace to query and investigate your Coralogix data without leaving your IDE. Learn more
Olly through MCP — investigate from your IDE in natural language
Five new MCP tools (olly_ask, olly_list_chats, olly_get_chat, olly_list_artifacts, olly_get_artifact) let you send investigation requests to Olly from Cursor, Claude Code, or Codex and get back answers plus downloadable artifacts (logs, traces, metrics, alert data, Kubernetes queries).
Olly is opt-in. Add the x-allow-costly-use: true header to your MCP server configuration to expose the tools — without it, they are not advertised. Olly tools are v2 only.
MCP server v2
A new MCP server version with a cleaner, AI-agent–optimized experience. To opt in, send the mcp-version: v2 header — see Setup. v1 is deprecated and will be removed.
- Sharper tool set — shorter tool descriptions and token-optimized outputs.
- Improved field discovery — semantic field search by description, plus value search through the new search-by-value API (
$d/$l/$mDataPrime paths). - Better schema and metric discovery — refreshed
get_schemas,search_metrics, andget_metric_details. - Stricter input validation — fewer silent failures and clearer errors.
Eleven new management tools across alerting, metrics, SLOs, TCO, Notification Center, and access control
New unified manage_* tools that let you create, read, update, and delete configurations across more of the Coralogix platform from your AI agent:
- Alerting —
manage_alerts_scheduler(alert suppression rules) andmanage_webhooks(outbound webhooks). - Metrics —
manage_recording_rules_groups_set(recording rules) andmanage_events2metrics(Events2Metrics). - SLOs —
manage_slo_v2s. - TCO Optimizer —
manage_tco_log_policiesandmanage_tco_trace_policies. - Notification Center —
manage_connectors,manage_presets, andmanage_global_routers. - Access control —
manage_custom_roles.
See the full list in the MCP tool catalog.
Alerts and parsing rules management
Alerts management via MCP — Create, read, update, and delete alerts across 12 alert types from your AI coding environment, and generate Terraform HCL, Kubernetes Operator YAML, and OpenAPI definitions for any alert. Learn more
Parsing rules management via MCP — Create, read, update, and delete parsing rules across 10 rule types, and generate Terraform HCL, Kubernetes Operator YAML, and OpenAPI definitions for any rule. Learn more
RUM tools for frontend and mobile monitoring
Real User Monitoring (RUM) tools for MCP Server. Query and analyze frontend and mobile application performance directly from your IDE — find top errors, track user interactions, monitor Web Vitals, and get SDK setup guidance without leaving your coding environment. Learn more
Natural language querying for observability data
Smarter AI Integrations with MCP Server — MCP Server brings natural language querying to your observability data outside Coralogix. Integrate with agents like Claude or Cursor to build intelligent, context-aware assistants powered by Coralogix data. Learn more