docs: delete SRE Agent stub pages, redirect to holmesgpt.dev

The SRE Agent menu was already dropped from the welcome page nav.
Now delete the three underlying pages — main-features (a stub),
getting-started (Quick Setup is now wizard-driven), and the
holmesgpt-docs redirect-stub — and point inbound references at
holmesgpt.dev or platform.robusta.dev.

- Update conf.py redirects: main-features.html, getting-started.html,
  holmesgpt-docs.html, plus the existing ai-analysis.html and
  holmesgpt/index.html — all now go to holmesgpt.dev
- Update inbound refs in metric-providers-external.rst,
  builtin-alert-enrichment.rst, and playbook-reference/index.rst to
  external links instead of internal :doc: references

https://claude.ai/code/session_019GQrc8wVq14if8f7F3pp4f

Author: claude

docs/conf.py

@@ -134,8 +134,11 @@
     "tutorials/alert-custom-prometheus.html": "/master/configuration/alertmanager-integration/alert-custom-prometheus.html",
     "catalog/triggers/prometheus.html": "/master/configuration/alertmanager-integration/index.html",
     "playbook-reference/prometheus-examples/alert-remediation.html": "/master/playbook-reference/automatic-remediation-examples/index.html",
-    "configuration/ai-analysis.html": "/master/configuration/holmesgpt/main-features.html",
-    "configuration/holmesgpt/index.html": "/master/configuration/holmesgpt/main-features.html",
+    "configuration/ai-analysis.html": "https://holmesgpt.dev/",
+    "configuration/holmesgpt/index.html": "https://holmesgpt.dev/",
+    "configuration/holmesgpt/main-features.html": "https://holmesgpt.dev/",
+    "configuration/holmesgpt/getting-started.html": "https://holmesgpt.dev/installation/?tab=robusta-helm-chart",
+    "configuration/holmesgpt/holmesgpt-docs.html": "https://holmesgpt.dev/",
     # AI Analysis pages redirects to holmesgpt.dev (docs have moved there)
     "configuration/holmesgpt/builtin_toolsets.html": "https://holmesgpt.dev/data-sources/builtin-toolsets/?tab=robusta-helm-chart",
     "configuration/holmesgpt/permissions.html": "https://holmesgpt.dev/data-sources/permissions/?tab=robusta-helm-chart",

docs/configuration/holmesgpt/getting-started.rst

@@ -185,5 +185,5 @@ Next Steps
 ----------
 
 - Configure :doc:`alert routing </notification-routing/index>`
-- Set up :doc:`AI-powered insights </configuration/holmesgpt/getting-started>`
+- `Set up AI-powered insights <https://platform.robusta.dev/signup>`_
 - Learn about :doc:`common configuration options <metric-providers>`

docs/configuration/holmesgpt/holmesgpt-docs.rst

@@ -7,7 +7,7 @@ Ever feel overwhelmed by alerts that lack context? Robusta enriches alerts autom
 
 .. note::
 
-   **Looking for automatic AI enrichment?** Check out :doc:`HolmesGPT </configuration/holmesgpt/main-features>` for zero-configuration AI-powered alert enrichment that automatically investigates alerts and provides root cause analysis.
+   **Looking for automatic AI enrichment?** Check out `HolmesGPT <https://holmesgpt.dev/>`_ for zero-configuration AI-powered alert enrichment that automatically investigates alerts and provides root cause analysis.
 
 Builtin Alert Enrichment
 *********************************

docs/configuration/holmesgpt/main-features.rst

@@ -18,7 +18,7 @@ Playbooks Basics
 
 Playbooks are deterministic rules for responding to alerts and unhealthy conditions in a Kubernetes cluster.
 
-Playbooks are recommended for advanced use cases. Most users should start with :doc:`SRE Agent </configuration/holmesgpt/main-features>` of alerts first, which requires far less configuration.
+Playbooks are recommended for advanced use cases. Most users should start with the `Robusta SRE Agent <https://platform.robusta.dev/signup>`_, which requires far less configuration.
 
 How Playbooks Work
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^