Manifold Security

Manifold is an agentic AI Detection and Response platform that gives security teams runtime visibility into what AI agents actually do on enterprise endpoints — the tools they call, the systems they touch and the actions they take. It targets the blind spot left by AI security tooling that only inspects prompts and model outputs at the inference layer.

Manifold Security builds an agentic AI Detection and Response (AIDR) platform for autonomous AI agents running on enterprise endpoints. Its founding argument is that the first generation of AI security tooling was built to inspect text — prompts in, model outputs out, at the inference layer — which leaves the part that actually causes damage invisible: the tools an agent invokes, the files and credentials it reads, the commands it executes and the external services it reaches. Manifold instead captures agent behaviour at runtime on the endpoint, establishes a behavioural baseline per agent, and flags anomalies the moment activity drifts, producing a live map of every agent in the environment along with its connections to MCP servers, databases and external systems. The company positions this squarely at coding agents — the scenario where a developer's laptop runs an agent with access to source code, credentials and CI/CD pipelines — and stresses agentless deployment in days without new gateways or proxy layers. It was founded by Neal Swaelens, Oleksandr Yaremchuk and Michael McKenna, the team behind Laiyer AI and its open-source LLM Guard project, which Protect AI acquired in January 2024. Manifold announced an $8 million seed round on 18 March 2026 led by Costanoa Ventures, with Cherry Ventures, Rain Capital and Modern Technical Fund participating alongside former Uber CSO Joe Sullivan and former Google DeepMind CISO Vijay Bolina. In April 2026 it launched Manifest, a free supply-chain intelligence platform that graphs how agent skills, plugins and MCP servers behave, indexing more than 100,000 assets.

Category

Governance & Security

Pricing

Freemium, Contact for pricing

Target Market

CISOs, Security Engineers, Detection and Response Teams, Platform Security Leads, CTOs

Deployment

Cloud-first

• ✓
  Agent discovery

  Enumerates every agent running in the environment along with the MCP servers, tools, resources and skills each one invokes at runtime.

• ✓
  Behavioural baselining

  Captures API calls, file access, command execution and external service interactions, then establishes a normal profile per agent to detect drift.

• ✓
  Risk and exposure analysis

  Surfaces risky agent behaviour such as unauthorised system access or improper data exposure, mapped against the systems each agent reaches.

• ✓
  Detection and response

  Real-time monitoring with anomaly flagging and intervention, so a misbehaving agent can be stopped rather than reviewed after the fact.

• ✓
  Agentless deployment

  Installs in days on existing endpoint infrastructure without new gateways or proxy layers, avoiding the latency cost of inline inspection.

• ✓
  Manifest supply-chain graphs

  Free platform building execution graphs of what a skill calls and environment graphs tracking authorship, similarity and cross-registry relationships across 100,000+ indexed assets.

• •
  Shadow AI agent inventory

  Security teams find out which agents developers are already running, what those agents connect to, and where the unmanaged risk sits.

• •
  Coding agent containment

  Monitor agents that hold repository, credential and pipeline access so an over-broad action is caught before it reaches production.

• •
  MCP server supply-chain review

  Use Manifest's execution and environment graphs to check a third-party skill or MCP server before approving it for internal use.

• •
  Incident investigation for agent activity

  Reconstruct precisely which tools an agent called and which systems it touched, evidence that prompt and output logs cannot supply.

• •
  Detecting malicious agent plugins

  Manifold's own published research has identified squatted plugin scopes and skills recruiting agents into crypto activity across public registries.

Manifold publishes no list pricing for the enterprise platform — evaluation runs through sales, and none of the launch coverage disclosed pricing models, per-endpoint rates or minimums. What is genuinely free is Manifest, the supply-chain intelligence platform launched in April 2026 as open access with an indexed database of more than 100,000 agent components, search and review. That gives a security team a real way to get value before any commercial conversation, which is unusual at this stage. Enterprise features extend coverage to additional components such as browser extensions and MCP servers and fold the intelligence into the runtime platform. As a March 2026 seed-stage company with $8 million raised, expect design-partner-style commercial terms rather than a published rate card.

This page was written from 5 sources, 3 on domains other than manifold.security.

1. 1.manifold.security — manifold.securityvendor
2. 2.manifold.security — manifold raises 8m seed fundingvendor
3. 3.siliconangle.com — manifold raises 8m secure autonomous ai agents enterprise en
4. 4.siliconangle.com — manifest platform manifold targets ai agent supply chain sec
5. 5.fintech.global — manifold raises 8m seed round to secure ai agents at runtime