v1.3.0#
07/08/2026
Attention
Reply CMP v1.3.0 is now available v1.3.0 brings tenant-owned provisioning to Reply CMP with BYOR runners, a governed module registry, exact-plan execution, and secure Terraform state handling. It also delivers production-ready monitoring, a visual Discovery map, richer AI report workflows, and a redesigned provisioning experience.
What’s Changed#
v1.3.0 makes provisioning more flexible and auditable while improving the way teams explore cloud estates, monitor providers, and refine AI-generated reports.
🚀 New Features#
Tenant-Owned Provisioning and BYOR#
Provisioning can now run through customer-owned Azure DevOps or GitHub Actions infrastructure.
Configure tenant-owned Azure DevOps pipelines or GitHub Actions workflows as provisioning runners
Use customer cloud workload identities and federated credentials instead of static Reply CMP API tokens
Enroll a tenant-specific callback identity with the
Provisioning.Runnerapplication roleStore runner credentials as tenant-owned Key Vault references
Support tenant-owned Azure Blob Terraform state with Azure AD authentication, versioning, soft delete, and lease locking
Keep the Terraform state key deterministic and scoped to the tenant, deployment, and operation
Governed Module Registry#
The provisioning registry now provides a controlled lifecycle for Terraform modules and versions.
Publish managed catalog item versions with immutable, content-addressed artifacts
Pin module versions and provider lock files before execution
Validate module schemas, provider sources, references, outputs, and dependency cycles during ingestion
Generate resource configuration forms from module schemas
Use resource input references to connect outputs from existing resources to new configurations
Keep catalog and deployment changes auditable through immutable projections and events
Exact-Plan Execution and Policy Guardrails#
Every provisioning operation now follows a controlled plan and apply contract.
Materialize a complete, operation-specific Terraform workspace
Verify workspace manifests, artifact hashes, provider topology, and provider checksums before execution
Run
terraform init,validate, andplanbefore policy evaluationApply only the saved, reviewed plan; replanning during apply is rejected
Require explicit policy overrides with a recorded reason when a plan is denied
Track operation claims, heartbeats, callbacks, state locks, and unknown outcomes
Provide recovery flows for verified state, restored state, and confirmed apply outcomes
Monitoring#
Monitoring has been reworked for production use across Azure, AWS, and GCP.
Validate queries against provider metric capabilities before execution
Return provider-specific failures with clearer diagnostics instead of ambiguous results
Improve dashboard creation, editing, import/export, presets, and preview workflows
Strengthen alert rule validation, scheduling, evaluation state transitions, and notifications
Add result budgets and safer handling for large monitoring responses
Discovery Map and Relationship Exploration#
Discovery now gives users a more practical way to explore infrastructure topology.
Add a visual inventory map with lazy loading for larger resource estates
Browse typed relationships between discovered resources and their related neighbours
Preserve provider-sourced and inferred relationship origins in the graph model
Improve ARM type-specific enrichment concurrency to reduce throttling responses
AI Report Refinement#
Chatbot users can now continue working with a generated report instead of starting over.
Retain report generation context across the chat session
Refine report goals, layout, and generated HTML through follow-up requests
Add
Last Yearas a report-generator date rangeImprove report prompts, template generation, and agent execution context
Keep report design changes and rollback context available during the workflow
🔧 Improvements#
Provisioning UX: redesigned deployment pages, module cards, registry panels, operation timelines, status chips, resource catalogs, previews, and schema forms
Resource creation: added clearer validation, metadata handling, reference selection, and navigation across provisioned resources
Monitoring UX: added provider connection views, time-series previews, dashboard presets, and more consistent widget configuration
Chatbot reliability: added retry handling and improved session/report context management
🐛 Bug Fixes#
Fixed monitoring queries that requested unsupported provider metrics
Fixed Azure ARM type-specific discovery failures caused by excessive concurrency
Fixed provider status rendering in the provisioning module preview
Improved handling of monitoring provider failures and alert evaluation transitions
Improved report date handling and follow-up generation behaviour
📚 Documentation#
Added the provisioning BYOR operations runbook, including runner identity, state protection, exact-plan execution, rollout, rollback, and incident procedures
Documented the governed module registry, provider locks, schema validation, and resource references
Updated provisioning, monitoring, Discovery, and AI report documentation for the v1.3.0 workflows