Product Updates

Starting August 18, 2026, a user's Evo data is limited to the Snyk Organizations they can read.

Until now, everyone with access to Evo saw every repository, asset, and issue in the Tenant. Evo now
reuses your existing Snyk Organization hierarchy, so there is nothing new to model and no migration to
run. Nobody loses access until an administrator changes their role.

What's new:

• Tenant Viewer now grants read access to the assets and issues in the Organizations the user belongs to, and to nothing outside them

• Tenant Member grants no Evo access, letting administrators withhold Evo without removing Snyk platform access

• Full-Tenant access continues for Tenant Admin and for the two roles ending in "with Evo access"

• Organization names appear alongside repositories, distinguishing the same repository imported into more than one Organization

Note for administrators:

Existing users keep their access, and Tenant Viewers gain scoped read access for the first time. New
users join as Tenant Member, which carries no Evo access, so each one needs a role change. Scoped
users are read only, and assets without an Organization stay hidden, which today means all Agent
Supply Chain Security and Continuous Offensive Security data.

Documentation: Access and authentication

This update is available on September 14, 2026. It improves coverage and precision across Snyk Code: template files are analyzed, Java framework and library support is extended, detection gaps are closed in six languages, and .gitignore no longer hides committed files from analysis.

Template files

Snyk Code analyzes template files and follows data from your application code into the template. Cross-site scripting that only becomes exploitable where the template renders its output is now reported, with the data flow shown from the application through to the template.

Supported engines: Jinja2, Razor, FreeMarker, EJS, Handlebars, Pug, Thymeleaf, Twig, Mustache and Velocity.

Java

• Apache CXF — the web services framework, including its HTTP transport.

• OkHttp — the HTTP client.

• Spring Security OAuth2 Client — OAuth2 and OpenID Connect authentication.

• Google OAuth Client and Google API Client — Google authentication and API access.

• MSAL4J — Microsoft identity platform authentication.

• Apache Commons Collections — collection utilities.

• Flyway — database migrations.

• JAXB — XML binding.

• SAP Commerce (Hybris) — FlexibleSearch queries, with parameter binding recognised as safe.

Coverage for these libraries is added or improved, resulting in improved data flow analysis.

Python

LangChain LiteLLM is recognised as a source of untrusted data.

Java 25

Java 25 source is analyzed.

Rule coverage improvements

Detection coverage is extended for Java, Kotlin, C#, Go, JavaScript and PHP — additional sources, sinks and unsafe API patterns, in particular for cryptography.

Expect additional true positives and fewer false positives.

Files matched by .gitignore

• Committed files matched by a .gitignore rule are analyzed.

• Secrets committed and later matched by a .gitignore rule are now detected. Git ignores only untracked files, so this brings Snyk Code and Snyk Secrets in line with Git.

• These are pre-existing issues in code you already committed.

• The change applies across all interfaces and cannot be disabled.

Required for CLI scans. Upgrade to Snyk CLI 1.1307.0 or later, available from 26 August. Earlier versions keep the previous behavior, so CLI results will differ from your SCM and web results. SCM imports and scans need no action.

What is unchanged

• Untracked files matched by .gitignore remain excluded from analysis.

• .snyk exclude patterns behave as before, and remain the way to exclude a path deliberately.

• Severity levels and rule identifiers do not change.

Learn more in the Snyk Code documentation.

Starting August 17, 2026, Evo by Snyk uses a rebuilt model risk score to evaluate the AI models discovered in your organization.

The previous Risk Index is replaced by a single score from 0–1,000 that combines attack success rate with the potential impact of a successful attack, tested against categories including prompt injection, data exfiltration, and insecure code generation. The result is a score that reflects both how easily a model can be compromised and how much damage that compromise could cause — not just whether an attack succeeded.

What's new:

• Impact-weighted scoring: Risk scores now factor in the real-world consequence of an attack succeeding, not only its likelihood.

• Four severity bands: Scores map to low (0–249), medium (250–499), high (500–749), and critical (750–1,000), so you can triage at a glance.

• Broader attack coverage: Testing spans a wider range of direct and indirect attack categories against each discovered model.

• Framework-mapped: Findings map to OWASP LLM Top 10, OWASP Agentic Security, MITRE ATLAS, and NIST AI 600-1/100-2e2025, so you can connect model risk to the compliance frameworks you already report against.

• Policy-ready from day one: Evo applies default policies across common attack categories automatically, and you can create custom policies scoped to specific attack categories or goals to match your organization's risk tolerance.

If you have policies built on the old Risk Index: those policies remain visible but no longer evaluate under the new scoring. Review and recreate any Risk Index-based policies against the new model risk score to keep enforcement active.

Read more in the Risk intelligence documentation.

The "What"

Snyk is introducing Snyk Assist — an AI-powered support and guidance assistant available directly inside the Snyk UI.

Snyk Assist answers product, setup, and troubleshooting questions in plain language, grounded in Snyk's own documentation, release notes, and support knowledge articles. It can also retrieve read-only information about your current organization and group, and open a support case without you leaving the conversation.

Assist is available from the top right of the screen on every page, and is always accessible.

• Ask questions in plain language: product concepts, account setup, integrations, and day-to-day administration — "How do I set up SSO?", "What permissions does Snyk need for my GitHub org?", "How do I configure ignore rules and approvals?". Answers are grounded in Snyk documentation and tailored to your role, plan tier, and licensed products.
  

• Self-serve troubleshooting: conversational diagnostics drawn from Snyk's support and knowledge articles, so common problems can be resolved without waiting on a ticket.
  

• Information about your account: issue summaries by severity, issue triage and single-issue detail, projects and import targets, collections and container images, scan settings, integrations and their connectivity status, members and their roles, and a project's current ignore rules — for your current organization and group, retrieved with your own permissions.
  

• Support without leaving the conversation: Assist can suggest and create a support case and return the case ID and link. When it detects a capability Snyk does not support today, it can capture that as a feature request.
  

• Vulnerability lookups: look up known vulnerabilities for a package, or a specific package version, against the Snyk Vulnerability Database.

Snyk Assist requires the new Snyk UI. It is available only to users who have the new navigation experience enabled. If you are still on the classic navigation, switch via the user account menu to use Assist.

Snyk Assist is available to Team, Ignite and Enterprise plans.

The "Why"

Today, when a user has a question or hits a problem in Snyk, they have to leave the product to resolve it — searching the documentation, asking internally, or opening a support ticket and waiting for a reply. That context switch is where onboarding stalls and where routine administration work slows down.

Snyk already runs an AI assistant in the Snyk support portal, and it has deflected roughly 85% of support cases year to date. Snyk Assist brings that same proven capability into the product itself, at the moment the question actually comes up rather than after the user has gone looking elsewhere.

Positioning it as a support and onboarding assistant — rather than a general-purpose chatbot — is deliberate: it sets an accurate expectation of what Assist will and will not do, and keeps its answers anchored to Snyk's own content rather than to open-ended generation.

The "How"

No changes to your data, your integrations, or your scan workflows.

Snyk Assist is read-only. It can read your existing data using your own permissions, and it never modifies anything. It cannot change settings, ignore issues, run scans, or modify projects — where an action is required, Assist explains how to do it in the UI or via the API and leaves the action to you.

Access follows your existing permissions. Assist sees only what you can already see, and it is scoped to the organization and group you are currently working in.

Snyk is rolling out a redesigned navigation experience!

The classic sidebar — which listed every page across Tenant, Group, and Organization scopes at once — is replaced by a single breadcrumb bar at the top that lets users switch scope in one click via per-level drop-downs.

The left sidebar is condensed to core areas only (Analytics, Inventory, Projects, Issues, Policies, Settings), with all settings-related pages consolidated into one context-aware Settings hub.

The update also adds a "Dark Mode" option (System/Light/Dark) under the user account menu. and a UI/Cmd/Ctrl+K command search to jump directly to any page or setting for easy orientation.

• Navigation & Scope Selector: one-click navigation bar; unifies scattered settings into one hub.

• Navigation Search (Cmd/Ctrl+K): lets users jump straight to a page/setting without menu-hunting, especially useful as sidebar is condensed.

• Dark Mode: standard user-requested accessibility/preference option, now consistent across every page - allowing users a continuous experience with other, dark-mode enabled tools.

No data, integration, or workflow changes — UI/navigation only.

Users can toggle freely between classic and new navigation via the user account menu.

If not sure about a location of a setting - just type the setting/feature element using "Navigation Search" (Cmd/Ctrl+K search).

New Navigation KB article

We are pleased to announce Snyk CLI release, v1.1306.4

This release contains fixes and minor improvements. To learn more beyond what is highlighted below, please reference the full release notes.

This update includes the following:

• Clearer error messages when the CLI cannot reach a configured proxy. Failures now report the proxy URL the CLI attempted to use, with credentials redacted, along with the specific error code SNYK-CLI-0028, making proxy misconfiguration quicker to diagnose in CI and behind corporate networks.

• An updated embedded runtime, moving Node.js from 22.22.2 to 22.23.2 and OpenSSL from 3.5.5 to 3.5.7, which brings in fixes for six high-severity CVEs plus additional OpenSSL security fixes.

• Fixed vulnerabilities:

  • SNYK-JS-JSYAML-18593780

  • CVE-2026-45447

  • CVE-2026-48618

  • CVE-2026-48933

  • CVE-2026-56846

  • CVE-2026-56848

  • CVE-2026-58043

If you have any questions, feel free to reach out to the Snyk support team. We encourage everyone to upgrade to the latest version to take advantage of these improvements.

We are pleased to announce Snyk CLI release, v1.1306.3

This release contains security fixes. To learn more beyond what is highlighted below, please reference the full release notes.

This update includes the following:

• Updates bundled dependencies to remediate known vulnerabilities, keeping the CLI current for teams with supply chain policies on the tools running in their pipelines.

• Fixed vulnerabilities:

  • SNYK-JS-SHESCAPE-18319522 (CVE-2026-14257)

  • CVE-2026-69152

If you have any questions, feel free to reach out to the Snyk support team. We encourage everyone to upgrade to the latest version to take advantage of these improvements.

Snyk API & Web now supports NTLM v2 authentication for Web targets. Security teams can configure NTLM credentials (username, password, domain, workstation) directly in the target settings, enabling authenticated scans of Windows-authenticated applications without requiring separate tools.

Many enterprise customers — especially in financial services, insurance, and government — run internal applications protected by NTLM authentication. Without this capability, these applications remained unscanned, creating security blind spots.

Security engineers can now scan Windows-authenticated internal applications that were previously inaccessible, eliminating coverage gaps. To configure NTLM authentication, navigate to Target Settings > Authentication tab, locate the Basic Auth / NTLM module, and select the NTLM option.

To learn more, visit NTLM in our user documentation.

Starting August 4, 2026, Snyk secrets scanner is generally available to help secure your entire SDLC against credential leaks.

Stolen credentials remain a leading initial access vector in security breaches, and the rise of AI-generated code increases the risk of exposing sensitive data. You can now use Snyk secrets scanner to detect hardcoded credentials, API keys, and tokens across your repositories and prevent them from reaching commits or being leaked.

Key capabilities include:

• Machine learning driven detection and prevention across your development surfaces.

• Secrets detection for both human-written and AI-generated code.

• Governance and finding management directly within the Snyk web UI.

For more information, please read the blog and Snyk Secrets documentation

Until now, a developer could suppress a Snyk Code finding on their own — directly in the Web UI or IDE — with no review step.

Suppression happened silently and unilaterally, leaving security teams without a gate on what got ignored.

Ignore Approval Workflow changes that.

An ignore is no longer applied automatically — it becomes a request that only takes effect after a reviewer approves it. Developers submit ignores from where they work; a reviewer approves or rejects from the web; both sides are notified of the outcome. Suppression moves from an ungoverned action to a reviewed, auditable one.

What you can do

• Submit ignore requests from the CLI, IDE, or API — a reason is required, and the ignore is created as Pending rather than applied.

• Review from the web — reviewers with the review permission approve or reject on the Ignore Requests page.

• Auto-approve your own — reviewers can bypass triage for ignores they create themselves.

• Stay notified — reviewers are emailed on new requests and requesters on the outcome.

• Control access with roles — read / create / edit / review / delete permissions; by default Collaborators request and Admins review, with custom roles supported.

• Audit everything — export ignore-request data via the audit-logs API.