RELATED 01
PostHog MCP Server
OFFICIAL SERVERPostHog's hosted MCP endpoint for analytics queries, feature flags, experiments, errors and more
41kGitHub stars
40popularity score
Coralogix MCP Server
Hosted MCP endpoint for querying and configuring a Coralogix observability account
Official · coralogix.com Remote (hosted) · Streamable HTTP DevOps and monitoring
Connect your client
Remote (hosted)
Add in one click
Opens the app, which asks before it adds the server.
Or add it by command or config
Cursor.cursor/mcp.json
{ "mcpServers" : { "coralogix-server" : { "url" : "https://api.eu2.coralogix.com/mgmt/api/v1/mcp" } } }Claude Codeterminal
claude mcp add --transport http coralogix-server https://api.eu2.coralogix.com/mgmt/api/v1/mcpCodexterminal, then run codex mcp login coralogix-server
codex mcp add coralogix-server --url "https://api.eu2.coralogix.com/mgmt/api/v1/mcp"What it does
Coralogix runs this MCP server itself as a remote streamable HTTP endpoint, one per Coralogix region. An assistant connected to it can run DataPrime queries over logs and spans, run PromQL against metrics, read the service map and service catalog, look through infrastructure resources, dashboards, saved views and cases, and check billable data usage. It is aimed at troubleshooting from inside an editor or coding agent: find the errors, trace them to a service, then fix the code.
It also writes. Unified manage_* tools create, update, delete and export alerts, parsing rules, Events2Metrics, recording rules, SLOs, TCO policies, Notification Center routers and presets, custom roles, dashboards and AI Center evaluations, with export as Terraform HCL, Kubernetes Operator YAML or OpenAPI definitions named for alerts and parsing rules.
The tool catalog lists 70 tools for the current v2 server. Six of them drive Olly, Coralogix's own observability agent; Coralogix treats those as costly and hides them unless the client sends the header x-allow-costly-use: true. The catalog says it is generated from the live server manifest, so the set changes over time. v2 is the default; the older v1 can still be selected with an mcp-version: v1 header, but Coralogix marks it deprecated and due for removal.
What it needs
A Coralogix account whose role carries the needed permissions: the logs, metrics and spans data-API read permissions to query, and read plus write config permissions for each resource a manage_* tool should change. The endpoint must match the account's region (us1, us2, us3, eu1, eu2, ap1, ap2, ap3 or gov1). Sign-in is by browser OAuth or a personal API key sent as a Bearer header. Codex needs codex mcp login coralogix-server to finish OAuth.
Limits
Queries run against archive storage only. Data held only in high-tier Frequent Search is not reachable, so recent-time log questions can come back empty unless an S3 archive is set up and receiving logs (Limitations page). Olly tools stay hidden without the x-allow-costly-use header. The v1 server is deprecated and will be removed. The Permissions page names manage_webhooks and manage_connectors, which do not appear in the current 70-tool catalog.
Who it suits
Teams already sending telemetry to Coralogix who want an assistant to investigate incidents and edit monitoring configuration without leaving the editor. Because it is hosted by Coralogix, there is nothing to install, but every request goes to Coralogix and acts with the signed-in user's permissions; anyone who wants a read-only connection should use a key or role without config write permissions. It is of no use without a Coralogix account, and accounts with no archive storage will see little log data.
Tools 60 of 70
ai_center_list_applications- Lists the team's AI Center applications and whether guardrails are on for each
ai_center_list_evaluations- Lists prebuilt and custom evaluations, for all apps or one application and subsystem
get_service_catalog_entities_data- Reads latency, throughput, error rate, health and saturation for service-catalog entities over a time window
get_service_catalog_entity_type_schema- Shows the columns and labels available for one service-catalog entity type
get_service_map_relations_graph- Returns the graph of calls between services from tracing data for a time range
list_service_catalog_entities- Lists the known entities, such as service names, of one type
list_service_catalog_entity_types- Lists the entity types the account has service-catalog data for
get_alerts_object- Fetches one alert definition, alert event or case by ID
manage_alerts- Creates, reads, updates, deletes and exports alert definitions
manage_alerts_scheduler- Manages alert suppression schedules
search_alert_definitions- Searches the team's alert definitions with filters
get_case_events- Reads the event timeline of a case, with filtering by event type and paging
manage_cases- Inspects or changes one or more cases
list_dashboard_folders- Lists dashboard folders
manage_dashboard_folders- Creates or deletes dashboard folders
manage_dashboards- Reads and changes Custom Dashboards
search_dashboard- Finds dashboards by meaning or ID and returns an indexed summary
search_queries- Searches queries saved in dashboards by meaning or field
get_system_dataset_info- Describes one system dataset and its schema
list_datasets- Lists system and user datasets that DataPrime can query
search_for_fields- Finds log or span field paths from a description or sample value, up to 5 at once
get_data_usage_capabilities- Shows the labels, units and request limits of the data usage API
query_data_usage- Reads billable ingestion volume in daily or hourly buckets
describe_dataprime_construct- Explains one DataPrime command or function
list_dataprime_commands- Lists DataPrime commands and functions
query_dataprime- Runs a DataPrime query over logs, spans or named datasets
fetch_doc- Returns one Coralogix docs page as markdown
search_docs- Searches the Coralogix docs by title or path
describe_infrastructure- Summarises which infrastructure the account monitors and how to filter it
get_infrastructure_resource_details- Reads one aspect of named infrastructure resources
list_infrastructure_resources- Lists infrastructure resources matching a filter, a page at a time
get_metric_label_values- Lists the values of one metric label
get_metric_labels- Lists the label names of one metric
manage_events2metrics- Manages Events2Metrics definitions
manage_recording_rules_groups_set- Manages recording rule group sets
query_promql_instant- Runs a PromQL instant query against Coralogix metrics
query_promql_range- Runs a PromQL range query against Coralogix metrics
search_labels- Finds label names containing a substring across all metrics
search_relevant_metrics- Finds metrics by description, name pattern or label
olly_ask- Sends a question to the Olly agent (hidden unless the costly-use header is sent)
olly_get_chat- Reads one Olly chat with its history (gated)
olly_list_chats- Lists the user's Olly chats (gated)
manage_ai_custom_evaluation- Creates or updates custom AI evaluation policies and links them to apps
manage_ai_evaluation- Creates, updates or deletes an evaluation on an AI application
manage_ai_model_pricing- Replaces the team's per-model token pricing overrides
manage_custom_roles- Manages custom roles
manage_data_enrichments- Manages data enrichments
manage_global_routers- Manages Notification Center global routers
manage_integrations- Manages integrations through the Integration Service
manage_parsing_rules- Creates, changes, deletes and exports parsing rules
manage_presets- Manages Notification Center presets (marked beta)
manage_quota_allocation_rule_set- Manages the team's quota allocation rule set
manage_slo_v2s- Manages SLO v2 resources
get_view- Fetches a saved log view by ID
list_views- Lists saved log views
manage_views- Creates, updates or deletes a saved view
manage_tco_log_policies- Manages TCO policies for logs
manage_tco_rum_policies- Manages TCO policies for RUM data
manage_tco_trace_policies- Manages TCO policies for traces
get_datetime- Returns the current UTC time for building query time bounds