Platform Agent

Oscilar MCP

Work in Oscilar without leaving your AI tools.

Request a Demo

See a Product Tour

Where it Helps

The Oscilar MCP connects your team’s own AI client, such as Claude Desktop, claude.ai, or Claude Code, directly to your Oscilar environment. Analysts and builders can explore workflows, draft changes, and test them by prompting in plain language. Every action runs as the signed-in user, and every change lands as a draft. Teams move faster on policy changes without retyping work between tools, and nothing goes live without a person deploying it.

Key Capabilities

Read Access to Your Environment

The agent lists and inspects cases, decisions, events, schemas, attributes, workflows, the rule library, templates, queues, and agent configs.

Plain-Language Workflow Explanations

Get a step-by-step explanation of any workflow, including its full graph and differences between versions, so inherited logic is no longer a black box.

Draft-Only Authoring

Create and edit workflow drafts, add and rewire steps, adjust conditions and thresholds, and build inputs and decisions. Live and production targets are rejected.

Runs as the Authenticated User

Sign-in uses your normal Oscilar login. There is no shared service account. The agent inherits the user’s existing role, and every action is attributed to that user in the audit trail.

Works with Your AI Client

Connect Claude Desktop, claude.ai, Claude Code, or another MCP-compatible client. Each connector reaches one environment, so sandbox and production stay separate. Deploying, activating, and deleting stay human-only in the UI.

Curated Tools, Not Raw API Access

Oscilar designs the tool set so the agent expresses intent (“route scores under 700 to manual review”) and the MCP handles orchestration and validation underneath.

How it Works

Connect Your AI Client

Add the Oscilar MCP as a custom connector using the URL in Settings > API Keys.

The agent authenticates through your normal Oscilar login, with your role and permissions.

Prompt in Plain Language

Ask the agent to explain a workflow, draft a change from a policy document, or send test events through a draft. It shows each rule that fired and each decision.

Review and Deploy in the UI

Open the draft on the canvas, review it, and deploy through the UI. Every action is logged to the user’s audit trail.

Popular Use Cases

Risk & Policy Teams

Turn a written policy into a draft workflow, change a threshold on a draft version, and test the result with synthetic events.

Compliance & Audit

Understand what is running, trace why a decision was made, review user access, and read the audit log. Every action is attributed to a real user.

Platform & Operations Admins

Set up queues, case statuses, saved searches, and QC checklists in bulk, and onboard new analysts through the agent instead of clicking through the UI.

Request a Demo

More Platform Agents

Test Case Creation

Auto-generated test cases for complete, fast workflow validation.

Rule Recommendation

Continuous, data-driven rule tuning to catch more fraud with fewer false positives.