MCP App Store
by KeenEthics
TrueWatch icon

TrueWatch MCP Server for ChatGPT

By Dorian from TrueWatch

Overview

Bring TrueWatch observability data into ChatGPT and Codex through the TrueWatch OWL MCP Server. Query logs, metrics, distributed traces, events, infrastructure, and real user monitoring (RUM) data using natural language. Discover available data sources and fields, compare trends, and correlate telemetry to investigate application performance, errors, and potential causes. Get evidence-based summaries with query scope and limitations. Available data and tools depend on your connected workspace and permissions. This plugin focuses on read-only query and analysis workflows.

How to connect

  1. ChatGPT: Open the TrueWatch listing in the ChatGPT app directory and choose Connect. ChatGPT listing TrueWatch exposes 48 tools here.

Tools

owl.account.workspace.same_org.list

ChatGPT
List account workspaces in the same organization as the current DF-API-KEY workspace, sorted by workspace id descending. Use it to discover candidate workspaces whose workspace_uuid can be passed as workspace_uuids to owl.data.same_org.trace.query, for example when reconstructing a trace that spans multiple workspaces in one organization. Supports workspace_uuids filtering and before_workspace_id marker pagination, and returns only workspace_id, workspace_uuid, workspace_name, and version_type. The workspace_id is for pagination only and must never be used for queries. This tool does not run any query and does not replace the final same-organization check performed by Kodo.

owl.apm.list

ChatGPT
APM discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid APM sources or fields before drafting an APM query. Use mode=field with source to limit fields to one APM source. The result is used to fill owl.data.query source or to help write query_text and select for APM queries. This tool remains backed by the APM API even though it is now grouped under the data category.

owl.billing.list

ChatGPT
Billing discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid cloud billing sources or fields before drafting a billing query. Use mode=field with source to limit fields to one billing source. The result is used to fill owl.data.query source or to help write query_text and select for billing queries.

owl.catalog.entity_get

ChatGPT
Get a single unified catalog entity by urn. Use owl.catalog.entity_query first when you need to find a valid urn.

owl.catalog.entity_query

ChatGPT
List unified catalog entities for a given entity_type. Use owl.catalog.entity_type_query first to get a valid entity_type. The returned urn values are the main inputs for owl.catalog.entity_get and owl.catalog.entity_topology_query.

owl.catalog.entity_topology_query

ChatGPT
Query topology related to a single unified catalog entity. Use owl.catalog.entity_query or owl.catalog.entity_get first to get the target urn.

owl.catalog.entity_type_query

ChatGPT
List unified catalog entity types. Use this first when you need a valid entity_type for owl.catalog.entity_query.

owl.data.same_org.trace.query

ChatGPT
Query one trace across multiple same-organization workspaces by trace_id, for reconstructing a request that spans several workspaces in one organization. First call owl.account.workspace.same_org.list to discover candidate workspaces and collect their workspace_uuid values, then pass those as workspace_uuids here; if workspace_uuids is empty or absent, only the current DF-API-KEY workspace is queried. The trace_id condition is always enforced by the backend, so do not add it yourself. The response data is the raw DQL query result structure and is not unpacked into flat item rows. The backend, validated by Kodo, performs the final same-organization reachability check; when the cross-workspace feature is disabled, only the current workspace is allowed and any non-current workspace_uuids cause a parameter error. Do not fall back to owl.data.query or owl.data.simple_query for cross-workspace trace lookups.

owl.data.show_dql_namespace

ChatGPT
List supported DQL namespaces and whether each namespace accepts an index. Use this first to choose the correct dql_namespace for owl.data.query. Then use if_index to decide whether owl.data.query should include index. After choosing the namespace, use discovery tools such as owl.field_schema.get, owl.log_index.list, owl.metric.list, owl.rum.list, owl.apm.list, owl.network.list, owl.profile.list, owl.logging.list, or owl.billing.list to gather valid fields, sources, metric tags, and indexes, then use owl.data.search_dql_docs and owl.data.check_dql before sending a hand-written query_text to owl.data.query.

owl.data.simple_query

ChatGPT
Execute a simple query without hand-writing full DQL. Provide namespace, time range, and any optional source, select_clause, where_clause, or group_by_clause you need, and the service will assemble the final query automatically. Prefer this tool when the query is straightforward and you want to reduce DQL hallucinations. Use owl.data.query only when you need full custom DQL or PromQL. index is only meaningful for namespaces L and R.

owl.errors.comment.add

ChatGPT
Add a comment to an error issue returned by owl.errors.list.

owl.errors.comment.list

ChatGPT
List comments for an error issue returned by owl.errors.list.

owl.errors.comment.update

ChatGPT
Update an existing error comment. This tool no longer handles comment creation; use owl.errors.comment.add for new comments.

owl.errors.list

ChatGPT
List all issues in the error center within a time range. Use this as the entry point for the errors domain before calling comment tools on a specific issue.

owl.event.get

ChatGPT
Get event detail by event document id, optionally scoped by a two-element timeRange.

owl.event.list

ChatGPT
List events in a required time range, with optional status and limit filters. Response items may include df_dimension_tags with event dimension tags.

owl.field_schema.get

ChatGPT
Field schema discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace when you need valid fields for filters, select clauses, or aggregations. It returns the current field directory or field catalog used to guide query writing, rather than a source-specific schema for a single datasource. This tool remains backed by the field schema API even though it is now grouped under the data category.

owl.incident.list

ChatGPT
List incidents with optional search and pagination.

owl.incident_comment.add

ChatGPT
Add an incident comment by incident UUID.

owl.incident_comment.list

ChatGPT
List incident comments by incident UUID.

owl.incident_operation.list

ChatGPT
List operation records for an incident.

owl.incident_schedule.get

ChatGPT
Get incident schedule detail by schedule UUID.

owl.incident_schedule.list

ChatGPT
List incident schedules with optional search.

owl.infrastructure.get

ChatGPT
Get infrastructure object detail by resource type and identity value. Use owl.infrastructure.list first when you need to locate the right identity_value.

owl.infrastructure.list

ChatGPT
List infrastructure objects such as hosts, containers, or processes. Use this first to find candidate objects before calling owl.infrastructure.get for a specific identity.

owl.llm.list

ChatGPT
List LLM applications with optional search, type, and pagination filters.

owl.log_index.get

ChatGPT
Log index detail helper for owl.data.query. Use this after owl.log_index.list only when you already have index_uuid and need more detail about a specific index. If you only need a valid index value for owl.data.query, prefer owl.log_index.list. This tool remains backed by the log index API even though it is now grouped under the data category.

owl.log_index.list

ChatGPT
Log index discovery helper for owl.data.query. Use this only when owl.data.query is targeting a log namespace and you need a valid index value before writing or executing the query. This tool remains backed by the log index API even though it is now grouped under the data category.

owl.logging.cluster_task.create

ChatGPT
Create an asynchronous logging cluster analysis task. Use this when log records need to be grouped by similar messages or fields. Provide query for a complete DQL statement, or provide namespace, conditions, timeRange, and analysisField so the backend can assemble the cluster query. The returned task_id must be passed to owl.logging.cluster_task.get, and callers should control the polling interval.

owl.logging.cluster_task.get

ChatGPT
Get the result of an asynchronous logging cluster analysis task. Call this after owl.logging.cluster_task.create returns task_id. task_status=ok means items and doc_count are ready; task_status=pending means the caller should poll again later.

owl.logging.list

ChatGPT
Logging discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid logging sources or fields before drafting a log query. Use mode=field with source to limit fields to one logging source. The result is used to fill owl.data.query source or to help write query_text and select for log queries.

owl.member.list

ChatGPT
List workspace members. The middleware auto-pages until all members are fetched.

owl.metric.list

ChatGPT
Metrics discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid metric sources, fields, or tag keys before drafting a metric query. Use mode=source for measurements, mode=field for field keys, and mode=tag for tag keys. When mode is field or tag, pass source to reduce noise to one metric source. This tool remains backed by the metric API even though it is now grouped under the data category.

owl.monitor.get

ChatGPT
Get monitor configuration by rule_uuid.

owl.monitor.list

ChatGPT
List monitors with optional name and status filters.

owl.monitor.receive

ChatGPT
Send an external event either to an existing receiver target, or by first creating an OuterEventChecker from monitor and then sending the event.

owl.monitor.upsert

ChatGPT
Create a new monitor when rule_uuid is omitted, or update an existing monitor when rule_uuid is provided.

owl.nbook_note.add

ChatGPT
Create a note for the current DF-API-KEY account and return its summary. Pass optional type as normal or runbook; if omitted, a normal note is created. Use owl.nbook_note.get with the returned note_uuid to retrieve the Markdown content.

owl.nbook_note.delete

ChatGPT
Delete a note by note_uuid. Returns note_uuid, path, and title for the deleted note.

owl.nbook_note.get

ChatGPT
Get a note by note_uuid. Returns the note details and Markdown content, including note_uuid, type, title, content, tags, tag_kv, is_public, is_favourited, created_source, creator_info, updator_info, create_at, and update_at.

owl.nbook_note.list

ChatGPT
List notes visible to the current DF-API-KEY account. Returns note summaries without Markdown content. Use note_uuid from a result with owl.nbook_note.get, owl.nbook_note.modify, or owl.nbook_note.delete. Pass optional type as normal or runbook to filter by note type; omit type to list all note types.

owl.nbook_note.modify

ChatGPT
Modify a note by note_uuid and return its updated summary. Provide at least one of title, content, tags, or tag_kv. Omit content to keep the current Markdown body. Note type cannot be changed after creation.

owl.network.list

ChatGPT
Network discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid network sources or fields before drafting a network query. Use mode=field with source to limit fields to one network source. The result is used to fill owl.data.query source or to help write query_text and select for network queries. This tool remains backed by the network API even though it is now grouped under the data category.

owl.pipeline.list

ChatGPT
List pipelines and return decoded pipeline content in a single call.

owl.pipeline.validate

ChatGPT
Validate and test a pipeline with the provided content and testData. Pass raw pipeline text and raw sample data directly; the service handles the internal encoding and test execution. Use this tool when you want to check whether a pipeline transforms sample data as expected.

owl.profile.list

ChatGPT
Profile discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid profiling sources or fields before drafting a profile query. Use mode=field with source to limit fields to one profiling source. The result is used to fill owl.data.query source or to help write query_text and select for profile queries. This tool remains backed by the profile API even though it is now grouped under the data category.

owl.rum.list

ChatGPT
RUM discovery helper for owl.data.query. Use this after owl.data.show_dql_namespace to discover valid RUM sources or fields before drafting a RUM query. Use mode=field with source to limit fields to one RUM source. The result is used to fill owl.data.query source or to help write query_text and select for RUM queries. This tool remains backed by the RUM API even though it is now grouped under the data category.

owl.slo.list

ChatGPT
List SLOs in the current workspace with optional name or monitor UUID search.

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/truewatch"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/truewatch)