Map / Runtime & Guardrails / CodeIntegrity

CodeIntegrity

USA · Founded 2024 · Raised $5.25M

Quick answer: CodeIntegrity is an AI security company based in the USA and founded in 2024, working in Runtime & Guardrails, MCP & LLM Gateways, Observability & Governance and 1 more category. It has raised $5.25M from investors including Syn Ventures, Antler and Boost VC.

Runtime & GuardrailsMCP & LLM GatewaysObservability & GovernanceAI Identity

Capabilities

Runtime control plane that sits between AI agents and enterprise systems, mediating every agent action before execution; turns agent instructions into sandboxed code so each action is explicit, inspectable, and repeatable — code becomes the control layer where intent, data, and tool calls can be checked before they run; full visibility into runtime tool actions with deterministic policy controls before execution; evaluates intent, data provenance, destination, and risk before any tool call executes; ties users, agents, and MCP clients to managed identity so every action has ownership; tracks the request, source context, destination, policy decision, and outcome behind each agent action for post-hoc security review; CodeIntegrity Labs tracks emerging agentic threats including prompt injection, tool abuse, and data leakage.

Sensors / integration

Runtime layer that inserts between agents and enterprise tools/systems; integrates with MCP clients; controls tool calls from agents and MCP clients before actions reach systems or leave approved boundaries; not a pure SDK — operates as a deterministic control plane that wraps the non-deterministic LLM in a code sandbox layer.

Protections

Forces unpredictable AI models to follow strict, deterministic rules — limits which systems and data an agent can touch; sandbox execution of agent-derived actions before they reach production systems; blocks prompt injection by separating the controlling model from the model that reads untrusted content; full audit trail tying users/agents/MCP clients to managed identity; enterprise boundary enforcement before action execution rather than after.

Differentiator

Made headlines compromising Notion AI in under 4 hours, earning Economist coverage in September 2025; among the first to articulate the "runtime control layer" concept as a category — sandboxing the model itself rather than just inspecting prompts; San Francisco-based team founded May 2024; direct competitors include Certiv, Raven, and Manifold — together forming the emerging "AARM" (Agent Action Runtime Mediation) cluster; deterministic-by-design philosophy contrasts with content-inspection-based runtime defense vendors.

Team & funding

Founders
Steven Jung, Abi Raghuram
Total raised
$5.25M
Investors
Syn Ventures, Antler, Boost VC

Other Runtime & Guardrails companies

AariaSec · ActaClad · ActiveFence · Acuvity · AGAT Software · AgentFox · AI Aramid · AI Prompt Shield · Aiceberg · Aim Security · AIR · Aira Security · See all →

Last updated · Get the “Listed on” badge