AI security posture management & discovery
Evo AI-SPM is an AI security posture management module from Snyk. It is a commercial product offered as: SaaS, CLI. Pricing model: usage-based.
Best for
Consider instead if
Evo AI-SPM is the posture module of Snyk Evo, and it starts where Snyk always has: in source code. It scans imported Go, Java, JavaScript, Python and TypeScript repositories, lists eight kinds of AI asset from models and datasets to MCP clients, and writes the result as a CycloneDX AI-BOM that the CLI and a REST API can also produce. The second half is a Model Risk Score: Snyk attacks each model repeatedly behind a basic system-prompt guardrail and publishes how each test goal maps to OWASP, MITRE ATLAS and NIST AI 600-1. Wiz reaches AI through cloud accounts; Evo reaches it through the code that is about to ship. The product page also claims running applications, but the docs we read describe repository discovery only. Evo needs an Enterprise contract.
This summary is our own wording of the sourced facts on this page.
Each value links to the page that states it, with the date we checked it.
As documented by the vendor or project. The tag is our label from a controlled vocabulary.
Our editorial model, not a standard. The primary layer follows the category, except for tools whose main documented job is detection and response; other layers rest on a capability the sources document. How the model works.
| Layer | Role | How it applies | Based on |
|---|---|---|---|
| L2 Discovery Discovery and posture | Primary | Find the AI models, services, agents, MCP servers and AI apps in use, and check how each is configured and exposed. | Follows its category, AI security posture management & discovery |
The risk list is OWASP's (Top 10 for LLM Applications, 2025). The mapping is our analysis, and each row rests on a capability the sources document.
| Risk | How it applies | Based on |
|---|---|---|
| LLM03 Supply Chain | Inventories models, datasets and MCP servers and writes an AI-BOM, which gives supply chain checks a starting list. | Continuously maps AI models, agents, frameworks, datasets, MCP servers and Skills across code and running applications; Generates a CycloneDX v1.6 AI-BOM for Python, Java, JavaScript and Go projects with the snyk aibom command; Detects eight AI asset types, from models and datasets to MCP clients and resources, in Go, Java, JavaScript, Python and TypeScript repos |
| LLM01 Prompt Injection | Snyk's mapping ties its role deviation test and every indirect injection goal to LLM01. | Scores each model across five security categories with an Attack Success Rate index and tests models in agent archetypes; Defines a 0 to 1,000 Model Risk Score from how often repeated adversarial attacks succeed and how severe the result is; Maps each test goal to MITRE ATLAS 5.6.0, OWASP LLM Top 10 2025, OWASP Top 10 for Agentic Applications 2026 and NIST AI 600-1 |
| LLM02 Sensitive Information Disclosure | Model risk tests include PII extraction, secret theft and private data leakage, which Snyk maps to LLM02. | Scores each model across five security categories with an Attack Success Rate index and tests models in agent archetypes; Defines a 0 to 1,000 Model Risk Score from how often repeated adversarial attacks succeed and how severe the result is; Maps each test goal to MITRE ATLAS 5.6.0, OWASP LLM Top 10 2025, OWASP Top 10 for Agentic Applications 2026 and NIST AI 600-1 |
| LLM06 Excessive Agency | Unauthorized tool use and destructive actions are test goals that Snyk maps to LLM06. | Scores each model across five security categories with an Attack Success Rate index and tests models in agent archetypes; Defines a 0 to 1,000 Model Risk Score from how often repeated adversarial attacks succeed and how severe the result is; Maps each test goal to MITRE ATLAS 5.6.0, OWASP LLM Top 10 2025, OWASP Top 10 for Agentic Applications 2026 and NIST AI 600-1 |
| LLM07 System Prompt Leakage | System prompt and tool extraction are test goals that Snyk maps to LLM07. | Scores each model across five security categories with an Attack Success Rate index and tests models in agent archetypes; Defines a 0 to 1,000 Model Risk Score from how often repeated adversarial attacks succeed and how severe the result is; Maps each test goal to MITRE ATLAS 5.6.0, OWASP LLM Top 10 2025, OWASP Top 10 for Agentic Applications 2026 and NIST AI 600-1 |
More in AI security posture management & discovery, ranked within the category by Coverage Score.
Finds AI workloads across Azure, AWS and Google Cloud and flags risks with recommendations and attack paths.
Best for: Teams already using Defender for Cloud that want AI workloads in Azure, AWS and Google Cloud added to their posture view
Reviewed
Google (Wiz)
Finds AI services across cloud accounts without agents and maps misconfigurations and attack paths.
Best for: Cloud security teams that want agentless discovery of AI services and their risks inside a cloud security platform
Reviewed
Noma Security
Finds agents, MCP servers, skills and models across endpoint, SaaS and cloud and flags their risks.
Best for: Security teams that need to find and govern agents, MCP servers and models spread across endpoints, SaaS and cloud
Reviewed
No. Snyk's plans page says Evo capabilities are not available on the Free or Team plan, and lists Evo AI-SPM on the Enterprise rate card.
An AI-BOM in CycloneDX v1.6 JSON for a local Python, Java, JavaScript or Go project, with an option to upload it to your Snyk Organization to view in Evo.
Not on the pages we opened. The product page, docs overview and plans page show no availability label, so confirm the current status with Snyk.
Snyk's docs define it as a number from 0 to 1,000 that combines how often an attack succeeds against the model with how serious the result is, and say it is not a percentage or failure rate. Snyk runs its attacks against each model behind a baseline system-prompt guardrail, so a score describes the model with that hardening, not your own prompt.
Evo AI-SPM is closed source (proprietary).
Source Source, checked October 21, 2026: snyk.io (opens in a new tab) Checked
Evo AI-SPM is offered as: SaaS, CLI.
Source Source, checked October 21, 2026: docs.snyk.io (opens in a new tab) Checked
Pricing model for Evo AI-SPM: Usage-based: Enterprise rate card: Evo AI-SPM at 0.66 credits per Active Contributor per day, with 1 credit = $1. Evo capabilities are not available on the Free or Team plan; Enterprise pricing is by contact sales..
Source Source, checked October 21, 2026: snyk.io (opens in a new tab) Checked
Status of Evo AI-SPM: Active.
Source Source, checked October 21, 2026: docs.snyk.io (opens in a new tab) Checked
We have no commercial relationship with Snyk.