Release Notes

OpenLM now speaks the Model Context Protocol (MCP), the open standard for connecting AI assistants to live business data. Point Claude, ChatGPT, Cursor, Windsurf, Gemini CLI, or any other MCP-aware client at your tenant, sign in once with OAuth, and ask questions in plain language: "Which features were denied most often last month?", "Show me underused AutoCAD seats by office." The OpenLM MCP Connector translates your prompt into a GraphQL query against your reporting database and returns tables, summaries, or — on higher-tier AI plans — fully interactive dashboards. No new BI tool to learn, no exported CSVs, no hand-built filters. Your reporting data, conversational.

Endpoints are provided for both regions — https://cloud-us.openlm.com/mcp for US and https://cloud-eu.openlm.com/mcp for EU. See the OpenLM MCP Connector documentation for client setup and the full tool reference.

LAC graduates from observation to enforcement. A new Agent enforcement engine prevents license consumption from workstations that are not running the Workstation Agent, bulk rule creation eliminates the per-rule call pattern that made large-option-file onboarding painful, and SaaS license servers join the supported targets for policy deployment. Deployment itself is more resilient, and the audit trail is finally complete. See the License Access Control changelog for the full per-version history.

• Agent enforcement (minimum viable product). LAC now correlates allocations against Agent Activity Manager to detect workstations consuming licenses without an active Workstation Agent. When the new global enforcement toggle is on, the next deployment skips allocations for those workstations — restoring accurate consumption data for high-value licenses and turning OpenLM from a passive observer into an active compliance control. Detection distinguishes a temporarily offline Agent from a missing one, so a brief disconnect does not punish legitimate users.
• Bulk allocation creation. Add hundreds of entities or features to a single asset in one action. Select multiple features and multiple entities at once in the allocation wizard, and LAC creates one allocation per combination — replacing the per-allocation pattern that previously made onboarding a 200-group option file an all-day task. Powered by a new AddRules GraphQL mutation; the existing AddRule mutation is unchanged.
• SaaS policy deployment. Policies can now be deployed to SaaS license servers through both scheduled and manual deployments, closing the gap between SaaS and on-premise coverage.
• Resilient deployment with corrupted Users & Groups Service (UGS) entities. Asset and policy deployments no longer fail when a referenced user or group has been disabled, deleted, or emptied in UGS. Affected allocations are skipped, logged with a clear warning, and surfaced in the deployment report, so administrators can clean up downstream without losing the rest of the deployment.

The new Homepage is a dashboard, not a lobby. Sign in and you see license health, denials, and pool utilization. The QuickSight page is gone, replaced by native widgets. The page loads faster and no longer needs the cloud. Every widget shares one shell, so loading, empty, and error states look and act the same. See the Homepage changelog for the full history.

• KPI summary cards for license servers offline and denied requests, each with a one-click deep link into the underlying view.
• License Servers Status donut breaking your fleet down into Healthy, Pending, and Error states, so a single outage no longer hides behind an aggregate.
• Top 5 Denied Features and Top 5 Features in Use, side by side — see where demand is hitting the ceiling and where engineering teams are spending the budget.
• Top 5 Saturated and Top 5 Underutilized License Pools, side by side — surface reclaim opportunities without writing a custom report.
• Usage trend and Upcoming expirations & renewals widgets, plus a severity-aware alert bar that surfaces critical signal at the top of the page.
• Take the Tour guided walkthrough for first-time admins, and a Software License Monitoring (SLM) activation gate that shows a clear lock card instead of empty widgets when SLM is inactive.

Updating Workstation Agents one machine at a time is over. From Agent Activity Manager, select any subset of agents across your fleet, pick a target Workstation Agent version, and trigger the upgrade in a single action. There is no per-machine MSI work, no need to touch endpoints individually, and rollout progress is visible in one place. Use it to deploy a hotfix to a single team, stage a phased rollout, or move an entire organization onto the latest agent on the same day.

LFM brings license-file editing, validation, and deployment into one place. Work safely with drafts before you push, see each file's parsed features as a structured table, compare versions at both the text and feature level, and let LFM keep license-file to license-server links in sync with SLM — triad-aware, with a full per-file event history. See the License File Management documentation for the full feature reference.

• Automatic synchronization between LFM and SLM, keeping license server names in sync with license files (including triad members).
• License file history with a per-file event timeline covering drafts, deployments, deactivations, and deletions.
• Parsed license features shown as a structured table — feature name, vendor, version, license type, start and expiration dates, quantity, and key.
• Compare different versions of the same license file: side-by-side raw-text comparison and a parsed-features table comparison that highlights added, removed, and changed features.
• Pre-validation of license files before pushing to Broker Hub, with file-text verification and warning detection (structural, semantic, and server-availability checks).

The License Parser — formerly a standalone OpenLM product — is now part of the OpenLM Platform. Drag a FlexLM license file onto it and get an instant, structured view of its contents: a Summary view, an Issued At view, and search across the parsed results. Files are parsed entirely in memory — nothing is written to storage — and the parser integrates directly with License Manager, with a clear message when an unsupported file type is uploaded.

Software Asset Management gains an AI-powered discovery suite. The Discovery Agent pipeline first discovers software vendors — headquarters, aliases, corporate websites — then discovers and enriches individual products, extracting vendor-defined feature codes (SKUs) and deployment types (on-premises, cloud, or hybrid). Successfully enriched products are promoted automatically to approved catalog entries, so your catalog stays authoritative without manual curation.

• Software Catalog Discovery page that orchestrates the discovery agents end to end.
• Discovered Vendors tab with a vendor grid, vendor management, and on-demand vendor discovery.
• Discovered Products tab with enrichment status and on-demand product discovery.
• Job Runs & History tab with active-job monitoring and a browsable history of every discovery run.

AI subscriptions are licenses too — and this release treats them that way. AI Proxy usage data is now aggregated hourly, the foundation for AI usage reporting. Identity resolution gets serious: Anthropic OAuth identities are resolved through the claude.ai profile endpoint and Admin API keys are recognized, so usage maps to people rather than anonymous keys.

A new Power BI dashboard turns that data into decisions across four report pages: an Executive Overview (total cost, total tokens, licensed users next to users with actual usage, top spenders, and cost by model family), Model Consumption Analysis, User Consumption Analysis — including each user's peak 5-hour token burst, the clearest signal for right-sizing license tiers — and an exportable day-by-day Details Report.