Command Zero
Also known as: Command Zero Throughline
Autonomous and AI assisted SOC platform whose Agent Zero takes alerts to a verdict around the clock, querying identity, endpoint, cloud, email and SaaS sources in place through read only APIs, from thousands of prebuilt and custom questions, with sign off on the verdict types the customer chooses, a replayable audit trail of every question and source, and a public API and MCP server.
Command Zero is an autonomous and AI assisted SOC platform that takes security alerts from alert to verdict across Tier 1 to Tier 3 work. Its agent, Agent Zero, investigates around the clock across identity, endpoint, cloud, email and SaaS data, querying each source directly at investigation time through read only APIs, with no ingestion pipeline, data migration or training data, and most environments go live in under an hour.
Command Zero, based in Austin with a presence in Calgary, was cofounded by chief executive Dov Yoran and chief technology officer Dean De Beer and took a $10 million strategic investment in July 2025 from Okta Ventures, SE Ventures and Crosspoint Capital. It reports more than 500,000 investigations completed and a 90 percent cut in Tier 1 escalations.
Investigations are built from questions. The platform ships with thousands of prebuilt questions drawn from SOC workflows and mapped to tools such as SharePoint, OneDrive, Exchange and Entra ID, and teams add their own, import detection logic from Splunk and CrowdStrike Next-Gen SIEM, and write questions for custom data sources. A knowledge base of enrichment data, analyst annotations and prior cases grows with each investigation, and Throughline merges related alerts into living investigations that reopen, rerun their questions and examine the verdict again as new alerts arrive.
The customer sets the policy. The SOC manager decides which verdict types need human sign off, which questions the agent may ask and which sources it may touch. Each verdict ships with the reasoning behind it, every question, source and artifact is recorded in order, past investigations can be replayed, and the audit trail belongs to the customer. Analysts can take over, redirect an investigation or ask for a verdict to be reconsidered, an Agent Validator reviews verdicts and flags edge cases, routine cases close themselves with documented reasoning, and incident response isolates hosts, disables accounts or blocks senders from templates.
A public API covers investigations, business context, catalog and schema and remediation, and an MCP server with 25 tools lets SOAR playbooks and assistants such as Claude start investigations and run remediation. The platform holds SOC 2 Type 2, names no model provider and offers no model choice, describes no single sign on or console roles, and is hosted by Command Zero with no region option. Gartner Peer Insights reviews come from retail, healthcare and education security teams. Licensing is based on the environment and the security team, quoted through sales.
Vendor details
Canonical URL
https://www.commandzero.ai
Category
Security / SOC agent
Funding status
Command Zero is independent and headquartered in Austin, Texas, with a presence in Calgary. It was co-founded by CEO Dov Yoran and CTO Dean De Beer, and it took a $10M strategic investment in July 2025 from Okta Ventures, SE Ventures and Crosspoint Capital (company release).
Company status
independent
Use cases & customers
Target customers
Deployment options
Integrations
Command Zero connects read only to 33 data sources across identity, endpoint, cloud, email and SaaS, including AWS (EC2 and IAM, S3, GuardDuty and EKS), CrowdStrike Falcon, Microsoft Defender for Endpoint, Microsoft 365, Entra ID, GitHub Enterprise, Splunk, Flashpoint and FortiDLP, and imports detection logic from Splunk and CrowdStrike Next-Gen SIEM. A public API and an MCP server with 25 tools let SOAR playbooks, custom hunting frameworks and assistants such as Claude start investigations, add business context and run remediation.
In practice
Your tier 2 and tier 3 work depends on a handful of people who know each tool deeply. Custom Questions encodes that logic into repeatable investigative sequences so any analyst or agent reaches the same conclusion.
You have just acquired a company and have no visibility into its environment. The federated read only model connects to its existing data sources with no ingestion or migration, so investigations run on day one.
You want investigations to run inside your own pipelines. The API endpoints and MCP server let SOAR playbooks, custom hunting frameworks and agents like Claude query the same governed, auditable data that produces every verdict.
Sources & related URLs
Research sources
Agentic Index coverage score
10.5 / 14 capabilities · 75%
| Integrations & Tool Calling | Full |
|---|---|
|
Command Zero connects read only to the customer's identity, endpoint, cloud, email, SaaS and SIEM tools, with Microsoft 365, Entra ID, Splunk and CrowdStrike Next-Gen SIEM named on the platform page and AWS, CrowdStrike Falcon, Microsoft Defender for Endpoint, GitHub Enterprise and Flashpoint among the sources named before. SOAR playbooks start investigations, and incident response isolates hosts, disables accounts or blocks senders from templates. Sourcecommandzero.ai/platformread 2026-11-19 |
|
| Workflow Orchestration | Full |
|
Agent Zero runs an investigation end to end from alert to verdict, choosing which of thousands of questions to ask across identity, endpoint, cloud, email and SaaS sources, chaining the answers and escalating the small set of alerts that need a person. People take part in the same run: analysts can take over, redirect or ask for a reconsidered verdict, and an Agent Validator, a human role, reviews verdicts and flags edge cases. Sourcecommandzero.ai/agent-zeroread 2026-11-19 |
|
| Knowledge Grounding & RAG | Full |
|
A knowledge base grows with every investigation as enrichment data, watchlists, analyst annotations and prior investigation context build up across cases, and Business Context APIs upload, list and retrieve the customer's own context. Teams add Custom Questions, import detection logic from Splunk and CrowdStrike Next-Gen SIEM, and write questions for custom data sources, so the investigative knowledge is the customer's as well as the platform's. Sourcecommandzero.ai/platformread 2026-11-19 |
|
| Human Oversight & Guardrails | Full |
|
The SOC manager decides which verdict types require human sign off, which questions Agent Zero may ask and which data sources it may touch. Each verdict ships with the reasoning behind it, so an analyst can confirm or override it, send it back for the agent to reconsider, or take over and redirect an investigation, and remediation runs from templates with an analyst's justification. Routine cases can close on their own with documented reasoning. Sourcecommandzero.ai/agent-zeroread 2026-11-19 |
|
| Security, Identity & Governance | Partial |
|
Command Zero holds SOC 2 Type 2 and calls its AI governed, with every step visible, auditable and reproducible under customer policy. No single sign on, role model or user permissions for the console are described on the site, so how access is divided inside a customer's own team is not shown, while the agent policy governs what Agent Zero itself may ask, touch and decide. Sourcecommandzero.ai/press/command-zero-raises-10m-to-scale-ai-driven-cybersecurity-earns-top-security-certificationread 2026-09-28 |
|
| Observability & Auditability | Full |
|
Every investigation records each question asked, data source queried and artifact considered, in order, every step is logged and every decision explained, and past investigations can be replayed so the result is reproducible. Every action the agent takes is logged with its inputs and outputs, and the audit trail belongs to the customer rather than to the platform that produced it. Sourcecommandzero.ai/platformread 2026-11-19 |
|
| Memory & State Persistence | Full |
|
Throughline merges related alerts into one living investigation with its evidence trail intact. When a new alert arrives the case reopens if closed, its time window extends, every question runs again and the verdict is examined afresh. That state persists, is scoped to the investigation and sits apart from the knowledge base, so the agent picks up a case where the evidence left it rather than starting over. Sourceprnewswire.com/news-releases/command-zero-introduces-throughline-living-investigations-debut-at-black-hat-usa-2026-302831167.htmlread 2026-09-28 |
|
| Deployment & Data Residency | Partial |
|
The data stays where the customer holds it: Command Zero queries endpoint, identity, cloud, email, SaaS and SIEM sources directly at investigation time through read only APIs, with no ingestion pipeline, data migration or parallel storage, and most environments go live in under an hour. The platform itself is hosted by Command Zero, and no region list, customer environment or on premises option is named. Sourcecommandzero.ai/platformread 2026-11-19 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
The platform ships with thousands of prebuilt investigative questions, built from SOC workflows and mapped to the tools they query, such as SharePoint, OneDrive, Exchange and Entra ID, along with remediation templates for isolating, disabling and blocking. Customers adopt them as they are and add their own, and packaged use cases cover alert triage, where routine cases close with documented reasoning, and incident response. Sourcecommandzero.ai/platformread 2026-11-19 |
|
| Triggers & Channel Coverage | Full |
|
Alerts start investigations automatically around the clock, with no scheduled downtime, related alerts reopen a Throughline investigation, and SOAR playbooks can start investigations through the API when alerts fire. Routine cases close themselves with documented reasoning, and the small set that needs a person is routed to an analyst with its supporting context attached. Sourcecommandzero.ai/agent-zeroread 2026-11-19 |
|
| Model Flexibility & Routing | Not documented |
|
No model provider, model choice or routing is named, and Agent Zero is described as a large language model orchestrating structured tool use under customer controlled policy. Claude and other assistants can reach the platform through the MCP server, but that is an outside assistant calling Command Zero, not a choice of the model that runs its own investigations. Sourcecommandzero.ai/agent-zeroread 2026-09-28 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
A public API covers investigations (list, start, extend, update, retrieve), business context (list, upload, retrieve), catalog and schema, and remediation (list templates, execute actions), plus an MCP server with 25 tools and slash commands that start investigations and run remediation, with scripts in a public GitHub recipes repository. How API and MCP access is authenticated is not stated, and the current site pages do not link the developer material. Sourcecommandzero.ai/blog/the-command-zero-api-and-mcp-server-are-liveread 2026-09-28 |
|
| Testing, Debugging & Optimization | Partial |
|
An Agent Validator, a human role, supervises Agent Zero's verdicts in production, flags edge cases and feeds mistakes back into retraining, and every verdict ships with its reasoning so a senior analyst can replay the chain and confirm or override it. That is review of live output. There is no customer facing harness, scored test set or comparison of agent versions before a change, and the performance figures are Command Zero's own. Sourcecommandzero.ai/agent-zeroread 2026-11-19 |
|
| Browser & Computer Use | Not documented |
|
There is no browser or computer use. Agent Zero queries data sources directly through read only APIs at investigation time, across identity, endpoint, cloud, email and SaaS tools, and response actions such as isolating a host, disabling an account or blocking a sender run through those tools' own interfaces rather than by an agent operating a screen or a browser session. Sourcecommandzero.ai/platformread 2026-09-28 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
Command Zero went live as a launch partner for Tenable OPEN for AI with an Exposure to Verdict investigation workflow. A priority exposure finding from Tenable feeds into a Command Zero investigation that checks whether the exposure was actually exploited, with the Tenable Hexa AI MCP Server and Command Zero's MCP server working in the same AI session. The approved verdict goes back to Tenable, and the workflow is published as a documented workflow and a ready to use skill on Tenable's CyberAgents Exchange.
Bears on: MCP / tool calling / API
View sourcePricing
Quote only, as an enterprise subscription
What is public
Command Zero sells a quote based enterprise subscription through sales, and evaluation also runs through sales, with no free tier or trial. Its Question Library, its API and MCP release and its SOC 2 Type 2 status are public, and its federated, read only architecture ingests and migrates no data.
Sales call required
Yes, required for paid access
Missing data
Command Zero does not state a rate, tiers, a billing unit, a minimum commitment or a cap, whether cost follows alerts, investigations, analyst seats or connected data sources, or whether API and MCP access are included or licensed separately.
Related vendors
- 7AI — Agentic SOC from the Cybereason founders: domain specialized agents…
- Abnormal AI — Behavioral AI email security with three named autonomous agents: a…
- AirMDR — AI native MDR whose AI analyst investigates every alert and writes…
- Airrived — Agentic OS that unifies SOC, GRC, IAM, vulnerability management, IT,…
- Andesite — Human-AI SOC platform where customers build their own agents and…
- Anvilogic — Agentic security operations platform that works on top of existing…
Alternatives to Command Zero
Other products that do the work Command Zero does, and what each one is.
- 7AICommand Zero vs 7AI →
- Dropzone AIAI SOC analystCommand Zero vs Dropzone AI →
- Prophet SecurityAI SOC platformCommand Zero vs Prophet Security →
- Radiant SecurityAI SOC platform triaging every alert typeCommand Zero vs Radiant Security →
- ReliaQuestAgentic security operations platform over the customer's existing stack
- CywareThreat intelligence and security operations platform