Provides tools for observability and configuration management for Dynatrace SaaS, including DQL queries, logs, spans, metrics, entities, problems, vulnerabilities, settings, dashboards, notebooks, SLOs, and synthetic monitors.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Dynatrace SaaS MCP Server list open problems with severity critical"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An MCP server for Dynatrace SaaS (Gen3 Platform) that exposes observability and configuration capabilities as tools for LLM agents (Claude Code, Claude Desktop, and any other MCP-compatible client).
This server bridges AI agents with Dynatrace SaaS, providing:
Observability — DQL queries, log search, span/trace lookup, metrics, entities, problems, vulnerabilities
Configuration — Settings 2.0 CRUD, dashboards, notebooks, SLOs, synthetic monitors
All write operations (create / update / delete) are blocked by default and must be explicitly enabled with DT_ENABLE_WRITES=true.
Related MCP server: MCP Datadog Server
Dynatrace SaaS exposes APIs across two distinct hostnames, each requiring a different token type.
Host | Env var | Example value | Auth header | APIs served |
Platform host |
|
|
|
|
Classic host |
|
|
|
|
Clone & install:
npm installConfigure environment:
cp .env.example .env
# then edit .env.env is gitignored — never commit it.
Variable | Required | Description |
| yes | Platform host base URL |
| yes | Classic host base URL |
| yes | Bearer token ( |
| yes | Api-Token ( |
| no | Set to |
| no | Request timeout in ms (default: 30000) |
dt0c01…)Needed for settings, metrics, entities, problems:
metrics.read
slo.read, slo.write
logs.read
settings.read, settings.write
entities.read
problems.read
ReadConfig, WriteConfig
ExternalSyntheticIntegration (synthetic read/write)
credentialVault.read
dt0s16…)Needed for DQL/Grail, document APIs, SLOs, synthetic, and vulnerabilities:
storage:logs:read
storage:metrics:read
storage:spans:read
storage:entities:read
storage:bizevents:read
storage:events:read
storage:buckets:read
storage:system:read
document:documents:read, document:documents:write
document:environment-shares:read, document:environment-shares:write
slo:slos:read, slo:slos:write
automation:workflows:read (optional, for synthetic execution)
securitySensor:vulnerabilities:read
# Install dependencies
npm install
# Build (compiles TypeScript to dist/)
npm run build
# Run the MCP server
node dist/index.js
# or equivalently:
npm start
# Development mode (tsx, no build step)
npm run dev
# Run tests
npm test
# Type-check only (no emit)
npm run typecheckAll mutating tools (create, update, delete) check the enableWrites config flag. If the flag is false (the default), calling any write tool returns an error immediately — no network request is made.
Set DT_ENABLE_WRITES=true in your .env (or in the MCP client env config) to enable writes. Even then, each destructive tool has its own description noting that it is a write operation.
Tool | Description |
| Execute a DQL statement against Grail |
| Validate a DQL statement without returning data |
| Search logs via DQL |
| Search spans/traces via DQL |
| Retrieve a trace by trace ID |
| Search metric keys (Grail) or list descriptors |
| Get metadata for a specific metric |
| Query metric data points |
| List host entities |
| Query entities with an entitySelector |
| Get details for a specific entity |
| List available entity types |
| List open or recent problems |
| Get details for a specific problem |
| List security vulnerabilities |
| Get details for a specific vulnerability |
Settings (8 tools): list_settings_schemas (filter with query), get_settings_schema, list_settings_objects, get_settings_object, validate_settings_object, create_settings_object, update_settings_object, delete_settings_object
OneAgent (1 tool): list_oneagent_features — exact feature keys, flags and objectIds from builtin:oneagent.features
Trace sampling (1 tool): get_trace_sampling_config — the trace ingest-control and URL/RPC sampling settings of one scope, in one call
Dashboards (5 tools): list_dashboards, get_dashboard, create_dashboard, update_dashboard, delete_dashboard
Notebooks (5 tools): list_notebooks, get_notebook, create_notebook, update_notebook, delete_notebook
SLOs (7 tools): list_slos, get_slo, evaluate_slo, list_objective_templates, create_slo, update_slo, delete_slo
Synthetic (7 tools): list_monitors, get_monitor, list_synthetic_locations, list_synthetic_nodes, create_monitor, update_monitor, delete_monitor
get_entity returns relationships and host properties inline, so it covers host-detail and entity-relationship use cases (there are no separate get_host / get_entity_relationships tools).
Synthetic location/node tools are named list_synthetic_locations / list_synthetic_nodes.
Add the following to your MCP client config (Claude Code ~/.claude/mcp.json or Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"dynatrace-saas": {
"command": "node",
"args": ["/Users/nasr/mycode/personal/saas-mcp/dist/index.js"],
"env": {
"DT_PLATFORM_URL": "https://asn8731h.sprint.apps.dynatracelabs.com",
"DT_CLASSIC_URL": "https://asn8731h.sprint.dynatracelabs.com",
"DT_PLATFORM_TOKEN": "dt0s16...",
"DT_API_TOKEN": "dt0c01...",
"DT_ENABLE_WRITES": "false"
}
}
}
}Replace the token values with real credentials. Set DT_ENABLE_WRITES to "true" only if you intend to allow the agent to create or modify Dynatrace configuration.
MIT — see LICENSE.
This project is MIT-licensed. It vendors one third-party component under its own license:
knowledge/dql/vendor/dynatrace-for-ai/dt-dql-essentials/ — the official Dynatrace
dt-dql-essentials skill from Dynatrace/dynatrace-for-ai,
redistributed unmodified under the Apache License 2.0
(see knowledge/dql/vendor/dynatrace-for-ai/LICENSE). Exposed via the dql_reference tool
(topic: "official" / officialRef).
knowledge/dashboards/vendor/dynatrace-for-ai/ — the official Dynatrace
dt-app-dashboards and dt-app-notebooks skills from Dynatrace/dynatrace-for-ai,
redistributed unmodified under the Apache License 2.0
(see knowledge/dashboards/vendor/dynatrace-for-ai/LICENSE). Exposed via the dashboard_reference tool
(tile types, visualizations, variables, full dashboard/notebook examples).
Add a user to an organisational level (IAM v1, WRITE). Requires an account-scoped platform token with iam write scopes. level-type is 'account' or 'environment'; level-id is the corresponding UUID.
Inputs: levelType (required), levelId (required), user (required).
Cancel a Grail record deletion process that is in progress or queued. Cancels subtasks that have not started yet; already-deleted data is NOT restored. Requires DT_ENABLE_WRITES=true. Required scopes: storage:records:delete, storage:events:read, storage:logs:read, storage:bizevents:read, storage:spans:read, storage:buckets:read.
Inputs: taskId (required).
Compare committed OpenAPI specs against the live tenant's catalog. Without args: returns a catalog-level overview of added/removed specs. With a spec stem (e.g. 'platform_slo_v1', 'environment-api-v2'): deep-diffs operations.
Inputs: spec.
Compare committed settings-schema versions against the live tenant. Optionally perform a structural key-path diff for a specific schemaId.
Inputs: schemaId.
Convert a legacy LQL (Log Query Language) matcher expression to a DQL (Dynatrace Query Language) equivalent (safe, read-only). Useful when migrating older pipeline routing conditions.
Inputs: body (required).
Create a new dashboard (WRITE). Sends a multipart/form-data POST per the Document Service spec. The content object is serialized to JSON and sent as the 'content' part with Content-Type application/json. Provide the content inline via 'content' OR from a file via 'contentPath' (not both). Requires DT_ENABLE_WRITES=true.
Inputs: name (required), content, contentPath.
Create an event-notification (WRITE, platform notification v2).
Inputs: notification (required).
Create a new monitoring configuration for an Extension 2.0 (WRITE). The config object must include at minimum a 'scope' field (e.g. HOST-XXXX) and a 'value' object.
Inputs: extensionName (required), config (required).
Create a feature in a project (WRITE). Requires feature-management:features:write scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), feature (required).
Create a flag in a feature (WRITE). Requires feature-management:features:write scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), featureKey (required), flag (required).
Create a Feature Management project (WRITE). Requires feature-management:projects:write scope and DT_ENABLE_WRITES.
Inputs: project (required).
Create a new filter segment (WRITE, platform storage filter-segments v1). Requires storage:filter-segments:write scope.
Inputs: filterSegment (required).
Create a new custom Grail retention bucket (WRITE, Dynatrace Storage Management v1). Requires DT_ENABLE_WRITES=true and the platform scope storage:bucket-definitions:write.
Inputs: bucket (required).
Create a group at an organisational level (IAM v1, WRITE). Requires an account-scoped platform token with iam write scopes.
Inputs: levelType (required), levelId (required), group (required).
Create a new synthetic monitor definition (WRITE). Pass a full monitor body per the platform synthetic v1 spec (SyntheticMultiProtocolMonitorRequest or SyntheticBrowserMonitorRequest).
Inputs: monitor (required).
Create a new notebook (WRITE). Sends a multipart/form-data POST per the Document Service spec. The content object is serialized to JSON and sent as the 'content' part with Content-Type application/json. Provide the content inline via 'content' OR from a file via 'contentPath' (not both). Requires DT_ENABLE_WRITES=true.
Inputs: name (required), content, contentPath.
Create a new request attribute (WRITE, classic Config API v1). Body must include at minimum: name, dataType, dataSources.
Inputs: requestAttribute (required).
Create a new request naming rule (WRITE, classic Config API v1). Body must include at minimum: naming, enabled, namingPattern, conditions.
Inputs: requestNaming (required).
Create a self-notification (WRITE, platform notification v1).
Inputs: notification (required).
Create a Settings 2.0 object (WRITE). Validate first with validate_settings_object.
Inputs: schemaId (required), scope (required), value (required).
Create an SLO (WRITE). Body fields per the SLO v1 spec (name, criteria, target, etc.).
Inputs: slo (required).
Create a WIF trust policy for an account (IAM v1, WRITE). Requires an account-scoped platform token.
Inputs: accountUuid (required), trustPolicy (required).
Create a WIF service user mapping for a trust policy (IAM v1, WRITE). Requires an account-scoped platform token.
Inputs: accountUuid (required), trustPolicyUuid (required), mapping (required).
Create a Dynatrace Automation workflow (WRITE, platform Automation v1). Requires DT_ENABLE_WRITES=true and automation:workflows:write scope on the platform token.
Inputs: workflow (required).
Delete (trash) a dashboard by id (WRITE). Optimistic locking: you must supply the current document version. The document is moved to the trash (restorable for 30 days). Requires DT_ENABLE_WRITES=true.
Inputs: id (required), version (required).
Delete an event-notification by ID (WRITE, destructive, platform notification v2).
Inputs: id (required).
Delete a specified monitoring configuration for an Extension 2.0 (WRITE, destructive).
Inputs: extensionName (required), configurationId (required).
Delete a feature by key within a project (WRITE, destructive). Requires feature-management:features:delete scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), featureKey (required).
Delete a feature flag by key (WRITE, destructive). Requires feature-management:features:delete scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), featureKey (required), flagKey (required).
Delete a Feature Management project by key (WRITE, destructive). Requires feature-management:projects:delete scope and DT_ENABLE_WRITES.
Inputs: projectKey (required).
Delete a filter segment by UID (WRITE, destructive, platform storage filter-segments v1). Requires storage:filter-segments:delete scope.
Inputs: uid (required).
Delete a Grail retention bucket by name (WRITE, destructive, Dynatrace Storage Management v1). This permanently removes all data in the bucket. The operation is asynchronous; the bucket status transitions to 'deleting' until complete. Cannot delete buckets whose name starts with 'default_' or 'dt_'. Requires DT_ENABLE_WRITES=true and the platform scope storage:bucket-definitions:write.
Inputs: bucketName (required).
Delete a synthetic monitor definition by entity ID (WRITE, destructive).
Inputs: id (required).
Delete (trash) a notebook by id (WRITE). Optimistic locking: you must supply the current document version. The document is moved to the trash (restorable for 30 days). Requires DT_ENABLE_WRITES=true.
Inputs: id (required), version (required).
Delete a request attribute by ID (WRITE, destructive, classic Config API v1).
Inputs: id (required).
Delete a request naming rule by ID (WRITE, destructive, classic Config API v1).
Inputs: id (required).
Delete a self-notification by ID (WRITE, destructive, platform notification v1).
Inputs: id (required).
Delete a Settings 2.0 object by objectId (WRITE, destructive).
Inputs: objectId (required).
Delete an SLO by id (WRITE, destructive).
Inputs: id (required).
Delete a WIF trust policy by UUID (IAM v1, WRITE, destructive). Requires an account-scoped platform token.
Inputs: accountUuid (required), trustPolicyUuid (required).
Delete a WIF service user mapping by UUID (IAM v1, WRITE, destructive). Requires an account-scoped platform token.
Inputs: accountUuid (required), trustPolicyUuid (required), mappingUuid (required).
Delete a Dynatrace Automation workflow by ID (WRITE, destructive, platform Automation v1). Requires DT_ENABLE_WRITES=true and automation:workflows:write scope on the platform token.
Inputs: id (required).
Evaluate an SLO by ID and return its compliance result (platform SLO v1). Uses the async evaluation:start / evaluation:poll flow per spec.
Inputs: id (required), timeframeFrom, timeframeTo.
Execute a Dynatrace Query Language (DQL) statement against Grail and return the result records. Use for logs, spans/traces, events, metrics, and entities. Example: 'fetch logs | filter loglevel == "ERROR" | limit 50'.
Inputs: query (required), maxResultRecords.
DESTRUCTIVE — permanently deletes Grail records matching the given DQL query and optional timeframe. Data that is deleted CANNOT be restored. Timeframe length is limited to 24 hours; end time must be at least 4 hours in the past. Subsequent requests are queued and executed in order. Returns a taskId to track deletion progress via get_record_deletion_status. Requires DT_ENABLE_WRITES=true. Required scopes: storage:records:delete, storage:events:read, storage:logs:read, storage:bizevents:read, storage:spans:read, storage:buckets:read. Allowed query commands: fetch, fields, fieldsAdd, fieldsRemove, filter, parse.
Inputs: query (required), timeFrame, timezone, locale.
Queries Smartscape entities via Grail DQL (Gen3-native). Pass useClassic:true to use the classic Entities v2 API on Gen2 tenants. The entitySelector must include a type(...) clause for the Grail DQL path; without it, a guidance message is returned.
Inputs: entitySelector (required), from, to, pageSize, +1 more.
Get a single audit log entry by its log ID (classic Audit Logs API, requires auditLogs.read scope). Uses GET /api/v2/auditlogs/{id}.
Inputs: id (required).
Get a single custom service definition by technology and ID (classic Config API v1, read-only).
Inputs: technology (required), id (required).
Get a dashboard by id — returns combined metadata and content. Fetches metadata from /documents/{id}/metadata and content from /documents/{id}/content, then merges them into { meta, content }.
Inputs: id (required).
Queries a single Smartscape entity by id via Grail DQL (Gen3-native). Pass useClassic:true to use the classic Entities v2 API on Gen2 tenants.
Inputs: entityId (required), useClassic.
Get a single event-notification by ID (platform notification v2).
Inputs: id (required).
List all versions of a specific Extension 2.0 by name.
Inputs: extensionName (required).
Get the environment-level configuration for an Extension 2.0.
Inputs: extensionName (required).
Get details of a specific monitoring configuration for an Extension 2.0.
Inputs: extensionName (required), configurationId (required).
Get a feature by key within a project (Dynatrace Feature Management API). Requires feature-management:features:read scope.
Inputs: projectKey (required), featureKey (required).
Get a feature flag by key (Dynatrace Feature Management API). Requires feature-management:features:read scope.
Inputs: projectKey (required), featureKey (required), flagKey (required).
Get a Feature Management project by key (Dynatrace Feature Management API).
Inputs: projectKey (required).
Get a filter segment by UID (platform storage filter-segments v1).
Inputs: uid (required).
Get a single Grail retention bucket definition by name (Dynatrace Storage Management v1).
Inputs: bucketName (required).
Get a single active user by UUID at an organisational level (IAM v1). Requires iam:users:read scope.
Inputs: levelType (required), levelId (required), uuid (required).
Get the descriptor/metadata for a single metric by key (classic Metrics v2). On Gen3/Grail tenants this endpoint may be unavailable; use query_metric or execute_dql instead.
Inputs: metricKey (required).
Get the full definition of a single synthetic monitor by its entity ID.
Inputs: id (required).
Get a notebook by id — returns combined metadata and content. Fetches metadata from /documents/{id}/metadata and content from /documents/{id}/content, then merges them into { meta, content }.
Inputs: id (required).
Get the full OpenPipeline configuration for a specific data type (id). Returns all endpoints, pipelines, and routing rules. Use this to inspect or fetch the config before modifying it with update_openpipeline_configuration.
Inputs: id (required).
Get the processors available for a specific OpenPipeline technology. Use list_openpipeline_technologies first to find the technology id.
Inputs: id (required).
Get full details of one problem, including root cause and affected entities.
Inputs: problemId (required).
Get the status of a previously submitted Grail record deletion process. POST /platform/storage/record/v1/delete:status with the taskId from execute_record_deletion. Returns one of: finished, submitted, processing, unknown, or failed, plus optional progress and message fields. Required scopes: storage:records:delete, storage:events:read, storage:logs:read, storage:bizevents:read, storage:spans:read, storage:buckets:read.
Inputs: taskId (required).
Get a single request attribute definition by ID (classic Config API v1).
Inputs: id (required).
Get a single request naming rule by ID (classic Config API v1).
Inputs: id (required).
Get a single self-notification by ID (platform notification v1).
Inputs: id (required).
Report this MCP server's build identity: version, git commit, build time, write-mode, and target Dynatrace hosts. Use to verify which build is running.
Inputs: no named parameters.
Get one Settings 2.0 object by objectId.
Inputs: objectId (required).
Get the full JSON schema for a Settings 2.0 schemaId. Use this to construct a valid 'value' before writing.
Inputs: schemaId (required).
Get one SLO by id (platform SLO v1).
Inputs: id (required).
Fetch all spans for a single trace id (ordered by start time) for latency/root-cause analysis.
Inputs: traceId (required), from.
Get a WIF trust policy by UUID, including its service user mappings (IAM v1).
Inputs: accountUuid (required), trustPolicyUuid (required).
Get full details of a single vulnerability by its ID (Dynatrace Platform Vulnerabilities v1).
Inputs: id (required), startTime, endTime.
Get a single Dynatrace Automation workflow by ID (platform Automation v1). Requires automation:workflows:read scope.
Inputs: id (required).
Get a single Dynatrace Automation workflow execution by ID (platform Automation v1). Requires automation:workflows:read scope.
Inputs: id (required).
Ingest a business event (CloudEvent) into Dynatrace via POST /api/v2/bizevents/ingest (WRITE). Requires DT_ENABLE_WRITES=true and a classic API token with the bizevents.ingest scope. Required CloudEvent fields: id, source, specversion, type. Optional: data, datacontenttype, dataschema, dtcontext, subject, time, traceparent.
Inputs: bizevent (required).
Ingest a custom event into Dynatrace via POST /api/v2/events/ingest (WRITE). Requires DT_ENABLE_WRITES=true and a classic API token with the events.ingest scope. Required fields: eventType (e.g. CUSTOM_INFO, CUSTOM_DEPLOYMENT, AVAILABILITY_EVENT), title. Optional: entitySelector, properties (key/value map), startTime, endTime, timeout.
Inputs: event (required).
Ingest one or more log records into Dynatrace via POST /api/v2/logs/ingest (WRITE). Requires DT_ENABLE_WRITES=true and a classic API token with the logs.ingest scope.
Inputs: logs (required).
List audit log entries (classic Audit Logs API, requires auditLogs.read token scope). Returns the auditLogs array. Supports time-range filtering via from/to and a free-text filter query. Uses the classic Environment API v2: GET /api/v2/auditlogs.
Inputs: filter, from, to, pageSize.
List all custom service definitions for a given technology (classic Config API v1, read-only).
Inputs: technology (required).
List Dynatrace dashboards accessible to you (Document Service). Filters to type='dashboard'. Optionally pass pageSize (max 1000).
Inputs: pageSize.
Queries Smartscape entity types via Grail DQL (Gen3-native). Returns guidance on common Grail entity tables. Pass useClassic:true to use the classic Entities v2 API on Gen2 tenants.
Inputs: pageSize, useClassic.
List event-notifications (platform notification v2). Returns paginated results array.
Inputs: no named parameters.
List all monitoring configurations for an Extension 2.0.
Inputs: extensionName (required), pageSize.
List all Extensions 2.0 available in the environment.
Inputs: pageSize.
List all flags in a feature (Dynatrace Feature Management API). Requires feature-management:features:read scope.
Inputs: projectKey (required), featureKey (required).
List all Feature Management projects (Dynatrace Feature Management API). Requires feature-management:projects:read scope.
Inputs: no named parameters.
List all features in a Feature Management project. Requires feature-management:features:read scope.
Inputs: projectKey (required).
List all Grail filter segments (platform storage filter-segments v1).
Inputs: no named parameters.
List all Grail retention bucket definitions (Dynatrace Storage Management v1). Returns the full bucket array including bucketName, table, retentionDays, status, etc.
Inputs: no named parameters.
Queries Smartscape host entities via Grail DQL (Gen3-native). Pass useClassic:true to use the classic Entities v2 API on Gen2 tenants. Optional tag/management-zone filters only work with useClassic:true (Gen2).
Inputs: tag, managementZone, pageSize, useClassic.
List visible groups at an organisational level (IAM v1). Requires iam:groups:read scope on an account-scoped platform token. level-type is 'account' or 'environment'; level-id is the corresponding UUID. Response array field: results (each item has uuid, groupName, type).
Inputs: levelType (required), levelId (required).
List active service users usable by the calling user at an organisational level (IAM v1). Requires iam:service-users:use scope.
Inputs: levelType (required), levelId (required).
List active users at an organisational level (IAM v1). Requires iam:users:read scope on an account-scoped platform token. level-type is 'account' or 'environment'; level-id is the corresponding UUID.
Inputs: levelType (required), levelId (required).
List/search metric descriptors (classic Metrics v2). Use a metricSelector like 'builtin:host.*' to filter. On Gen3/Grail tenants this endpoint may be unavailable; use query_metric or execute_dql instead.
Inputs: selector, pageSize.
List all synthetic monitors (browser and network availability). Optionally filter using a monitor-selector expression.
Inputs: monitorSelector.
List Dynatrace notebooks accessible to you (Document Service). Filters to type='notebook'. Optionally pass pageSize (max 1000).
Inputs: pageSize.
List SLO objective templates (platform SLO v1).
Inputs: no named parameters.
List all OpenPipeline configurations (one per data type: logs, events, bizevents, metrics, spans, davis, sdlcEvents, securityEvents, etc.). Returns the available data-type configurations with their editable status.
Inputs: no named parameters.
List, per data type, the processor types allowed in each pipeline stage (parsed from each configuration's pipelinesSpecification). Useful for discovering which processor types (e.g. 'fieldsAdd', 'dql', 'drop', 'geoLookup') are valid in a given stage for a given data type. Read-only.
Inputs: configId.
List all available OpenPipeline technology parsers (grouped by technology category). Each technology entry includes its id, name, matcher condition, and allowed configuration types.
Inputs: no named parameters.
List problems (classic Problems v2). Filter by status/severity via problemSelector and a timeframe.
Inputs: problemSelector, from, to, pageSize.
List all release stages in a Feature Management project. Requires feature-management:projects:read scope.
Inputs: projectKey (required).
List all request attribute definitions (classic Config API v1).
Inputs: no named parameters.
List all request naming rules (classic Config API v1).
Inputs: no named parameters.
List self-notifications (platform notification v1). Returns paginated results array.
Inputs: no named parameters.
List Settings 2.0 objects, filtered by schema and/or scope.
Inputs: schemaIds, scopes, pageSize.
List Settings 2.0 schema ids (classic). These identify configurable settings types.
Inputs: pageSize.
List Service-Level Objectives (platform SLO v1).
Inputs: pageSize.
List all synthetic locations (both public and private) available for the environment.
Inputs: no named parameters.
List all synthetic nodes (ActiveGates capable of running synthetic monitors).
Inputs: no named parameters.
List WIF (Workload Identity Federation) trust policies for an account (IAM v1). Requires an account-scoped platform token. Response array field: results.
Inputs: accountUuid (required).
List WIF service user mappings for a trust policy (IAM v1). Response array field: results.
Inputs: accountUuid (required), trustPolicyUuid (required).
List security vulnerabilities (Dynatrace Platform Vulnerabilities v1). Supports optional DQL-style filter expressions, sort, and time range.
Inputs: filter, sort, startTime, endTime.
List Dynatrace Automation workflow executions (platform Automation v1). Draft and simple workflow executions are not included. Requires automation:workflows:read scope on the platform token.
Inputs: workflow, state, limit, offset.
List Dynatrace Automation (Workflows) workflows via the platform Automation v1 API. Requires automation:workflows:read scope on the platform token.
Inputs: limit, offset, search.
Get DQL autocomplete suggestions for a DQL processor script (safe, read-only). Useful when authoring pipeline DQL expressions.
Inputs: body (required).
Get autocomplete suggestions for a matcher (routing condition) expression (safe, read-only). Useful when authoring pipeline routing conditions.
Inputs: body (required).
Preview the effect of an ORDERED SEQUENCE of processors by threading a record through each one via repeated /preview/processor API calls (safe, read-only). Each step's output record becomes the next step's input — this is pure-code orchestration, not a bulk API call. Returns a per-step trace (matched, record) and the finalRecord after all steps. Stops at the first step that returns an error.
Inputs: processors (required), sampleData (required).
Preview the effect of a pipeline processor on sample data without mutating any configuration (safe, read-only). Returns per-record match and transformed record results. Use this to author and validate processors before applying them with update_openpipeline_configuration. Include a 'sampleData' string field (JSON-encoded record) inside the processor definition.
Inputs: processor (required).
Query metric data points via Grail DQL 'timeseries' (Gen3-native, platform token). Builds a timeseries query and executes it against the Grail storage API. Works on all Gen3/Grail tenants with the storage:metrics:read scope.
Inputs: metricKey (required), aggregation, by, filter, +3 more.
Trigger a run of a Dynatrace Automation workflow (WRITE, platform Automation v1). Creates a new Execution for the workflow. Requires DT_ENABLE_WRITES=true and automation:workflows:run scope on the platform token.
Inputs: id (required), input.
Search logs in Grail via DQL. Builds a 'fetch logs' query with optional filters. For advanced needs use execute_dql.
Inputs: contains, loglevel, host, from, +1 more.
Search distributed-tracing spans in Grail via DQL ('fetch spans'). Filter by service, status, or duration.
Inputs: service, onlyErrors, minDurationMs, from, +1 more.
Send an email via the Dynatrace Email API (platform email v1). Requires DT_ENABLE_WRITES=true and the platform scope email:emails:send. The sender is fixed to no-reply@dev.apps.dynatracelabs.com.
Inputs: email (required).
Update an existing dashboard (WRITE). Uses PATCH /documents/{id} with multipart/form-data per the spec. Optimistic locking: you must supply the current document version. At least one of name, content, or contentPath must be provided. Requires DT_ENABLE_WRITES=true.
Inputs: id (required), version (required), name, content, +1 more.
Update an event-notification by ID (WRITE, platform notification v2).
Inputs: id (required), notification (required).
Update a specified monitoring configuration for an Extension 2.0 (WRITE). Scope cannot be changed; only the 'value' field is updated.
Inputs: extensionName (required), configurationId (required), config (required).
Update a feature by key within a project (WRITE). Requires feature-management:features:write scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), featureKey (required), feature (required).
Update a feature flag by key (WRITE). Requires feature-management:features:write scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), featureKey (required), flagKey (required), flag (required).
Update a Feature Management project by key (WRITE). Requires feature-management:projects:write scope and DT_ENABLE_WRITES.
Inputs: projectKey (required), project (required).
Update (replace) a filter segment by UID (WRITE, platform storage filter-segments v1). All fields are overwritten. Requires storage:filter-segments:write scope.
Inputs: uid (required), filterSegment (required).
Update a Grail retention bucket by name (WRITE, Dynatrace Storage Management v1). Uses an optimistic-locking PUT; include the current 'version' in the bucket body. Requires DT_ENABLE_WRITES=true and the platform scope storage:bucket-definitions:write.
Inputs: bucketName (required), bucket (required).
Update (replace) a synthetic monitor definition by entity ID (WRITE). Sends a full PUT to /monitors/{monitor-id} per the platform synthetic v1 spec.
Inputs: id (required), monitor (required).
Update an existing notebook (WRITE). Uses PATCH /documents/{id} with multipart/form-data per the spec. Optimistic locking: you must supply the current document version. At least one of name, content, or contentPath must be provided. Requires DT_ENABLE_WRITES=true.
Inputs: id (required), version (required), name, content, +1 more.
Replace the full OpenPipeline configuration for a specific data type (WRITE, requires DT_ENABLE_WRITES=true). Fetch the current configuration first with get_openpipeline_configuration, modify it, then PUT it back. This replaces the entire configuration object.
Inputs: id (required), configuration (required).
Update an existing request attribute by ID (WRITE, classic Config API v1).
Inputs: id (required), requestAttribute (required).
Update an existing request naming rule by ID (WRITE, classic Config API v1).
Inputs: id (required), requestNaming (required).
Update a self-notification by ID (WRITE, platform notification v1).
Inputs: id (required), notification (required).
Update an existing Settings 2.0 object by objectId (WRITE).
Inputs: objectId (required), value (required).
Update an SLO by id (WRITE).
Inputs: id (required), slo (required).
Update a WIF trust policy by UUID (IAM v1, WRITE). Requires an account-scoped platform token.
Inputs: accountUuid (required), trustPolicyUuid (required), trustPolicy (required).
Update (replace) a Dynatrace Automation workflow by ID (WRITE, platform Automation v1). Requires DT_ENABLE_WRITES=true and automation:workflows:write scope on the platform token.
Inputs: id (required), workflow (required).
Validate a settings value against the live schema (read-only, validateOnly). Returns violations, required-missing keys, and a corrected value filtered to known properties.
Inputs: schemaId (required), value (required), scope.
Validate a Settings 2.0 object payload WITHOUT persisting it (validateOnly=true). Returns constraint violations if invalid. Always safe (read-only).
Inputs: schemaId (required), scope (required), value (required).
Validate a DQL statement without returning data (executes with limit 0). Returns ok=true or the validation error.
Inputs: query (required).
Validate a DQL processor script without mutating any configuration (safe, read-only). Returns validation errors or a success indicator. Use this to author and validate DQL processing scripts before applying them with update_openpipeline_configuration.
Inputs: body (required).
Validate a matcher (routing condition) expression without mutating any configuration (safe, read-only). Use this to validate routing conditions before applying them with update_openpipeline_configuration.
Inputs: body (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 152 tools
Most tools target a distinct resource and action, with clear descriptions naming API versions and scopes. Some confusion is possible between generic execute_dql and purpose-built search_logs/search_spans/query_metric, and between v1/v2 notification tools, but these boundaries are mostly explained.
The vast majority follow a consistent snake_case verb_noun pattern (list/get/create/update/delete/ingest/execute). Minor deviations include get_extension actually listing versions, non-verb-noun names like check_api_spec_drift, and mixing verb families (validate/verify/check/preview) for similar safety operations.
152 tools is an extreme count for a single MCP server, far beyond the 25+ threshold for 'too many'. Even for a broad platform like Dynatrace, this creates a heavy selection burden and suggests the server is a dumping ground of many APIs rather than a coherent toolset.
Many resources have full CRUD (dashboards, SLOs, settings, monitors, workflows, buckets, feature flags, IAM). However there are notable dead ends: problems and vulnerabilities are read-only, custom services only list/get, and extensions lack create/delete for the extension itself.
Query your app's logs, metrics and incidents from your AI assistant.
Manage uptime monitors and incidents and search logs, metrics and traces from an AI assistant.
Open-source agent that observes and fixes your application. Query logs, traces, metrics, incidents.
LLM Orchestration Agent (Opentelemetry Api)
