Instructions + Tools.
Give it a job and the tools it needs. The simplest way to build.
Instructions
Tool selection
Settings
think → act → check → repeat.
Build Your AI Workforce
Build agents that fit your SOC.
Build your own agents your way, from a simple set of instructions to a full step-by-step workflow, and run them under the same governance as Imperum’s ready-made ones.
Ways to build
Build agents at the level of control you need.
Pick the level of control that fits the job, they all run on the same trusted foundation.
Step-by-step workflow.
Stages, branches, and approval points along the way.
Take in → Look up → Analyze → Decide → RespondBranches based on the outcome
Human approval points
Built-in history and context at every step.
Visual builder.
Drag-and-drop for workflows that go beyond a straight line.
Loops, branches, and moreLoops, parallel work, reusable pieces
Memory & double-checking
Runs reliably, with progress saved as it goes.
All three ways to build share the same governed access, full record, and approval rules.And you can start from any of the 500+ ready-made templates and make it your own.
Template Catalog
Don’t start from scratch.
500+ ready-made templates across 16 categories, start from any one and make it your own.
Months of SOC engineering, ready on day one.
Each template is a proven response pattern, triage, phishing, ransomware, lateral-movement hunts, cloud posture, audit evidence. The scoping and validation is done.
Start from the closest match, connect your tools, set your approvals, and go live. A quarter’s work, done in an afternoon.
500+ templates · 16 categories · build on any one, your way.
Alert TriageIncident ResponseThreat HuntingComplianceCloud Security+11
Alert Triage
Alert triage template
Structured
Incident Response
Phishing response template
Advanced
Cloud Security
Cloud-misconfig hunt template
Structured
Threat Hunting
Lateral-movement hunt template
Advanced
Compliance
NIST evidence-collection template
Structured
Incident Response
Endpoint-isolation playbook template
Advanced
Settings · Governance
Every agent runs under the same guardrails.
Every agent you build automatically follows the same rules as the rest of the platform.
PER-AGENT SETTINGS
InstructionsPROMPT
The agent’s role, in plain English, saved with each version.
Tool selectionTOOLS
Choose the tools the agent can use, each one governed by the gateway.
History & contextSCOPE
Limit what past data and context the agent can draw on.
Risk policyRISK
Mark each action as low, medium, high, or critical risk.
Approval rulesAPPROVALS
Set who approves which actions, and how long they have to respond.
AI modelMODEL
Choose which AI model the agent uses, on your servers or in the cloud.
Auto-approve limitAUTO
Decide how much the agent can do on its own before it needs sign-off.
Effort limitCEILING
A hard limit on how long the agent can keep working, so it never runs away.
TestingEVAL
Built-in tests to make sure the agent keeps performing as expected.
BUILT-IN GOVERNANCE
MCP Gateway
Every tool the agent uses goes through one secure, governed, fully logged gateway.
Approvals
High-impact actions pause for human approval before they run.
Privacy protection
Sensitive personal data is hidden before the AI sees it, and restored afterward.
Your history
Every agent draws on your past cases and knowledge to stay grounded.
Connected context
The agent understands how people, systems, and events connect.
Audit Log
Every action an agent takes is recorded in full and kept private to each customer.
Bring your runbooks. Ship them as agents.
Turn your team’s playbooks into working agents, running under the same governance as Imperum’s ready-made ones.
Imperum runs on your own servers, in the cloud, or fully isolated. Build your agents once, run them anywhere. Your data stays yours.