New agents and products for security, productivity, and creative work

Anthropic added new partner agents and products to Claude Marketplace across security, developer infrastructure, creative, and productivity domains, according to Claude Blog's report. The expansion brings the marketplace catalog to more than 25 partner solutions, allowing eligible enterprise customers to draw down their committed Anthropic spend directly against third-party integrations. This single-commitment billing structure enables enterprise accounts to consolidate payments whether purchasing new tools or expanding existing contracts alongside their core Claude deployments.

This breakdown details the security-oriented and operational tools integrated into the catalog, covering how technical teams deploy, monitor, and govern autonomous workflows. The piece serves engineering leads, enterprise security architects, and operations managers who need to evaluate automated remediation, repository vulnerability scanning, and agent access controls across enterprise environments. Teams evaluating automated runtime defenses and development pipelines can review how partner systems interact with source code, corporate knowledge bases, and endpoint fleets.

New agents and products for security: what actually changed

Anthropic broadened Claude Marketplace to include dedicated third-party tools focused directly on securing code, agent deployments, and cloud-connected infrastructure. The newly cataloged security offerings feature Chainguard Harness, Checkmarx One, Cyera Agent Guardian, Rubrik Code Guardian, and Tanium Cloud. Alongside these protective tools, the marketplace gained developer infrastructure from Datadog and Netlify, marketing software from AirOps, Runway, and Webflow, and productivity systems from Asana, Glean, monday, and Zoom.

The commercial structure of the marketplace now permits eligible enterprise customers to allocate a portion of their committed Anthropic software spend toward partner products. Instead of managing separate enterprise procurement streams and purchase orders for every external agent tool, companies can pool spend commitments under a single commercial umbrella. The platform now lists more than 25 integrated products and partner agents that enterprise organizations can browse and request directly through individual product directory pages.

New agents and products for security: how it works

The security products operate across distinct phases of the software development and deployment lifecycle. Chainguard Harness compiles applications directly from source code inside an isolated microVM, probing the executing runtime for exploitable weaknesses and generating custom remediation guidance for reproducible flaws that Claude discovers. Checkmarx One merges static rules-based code analysis with artificial intelligence reasoning to detect source vulnerabilities, helping developers prioritize and patch issues prior to production builds. Cyera Agent Guardian catalogs active Claude deployments, governs permissions prior to agent execution, enforces security policies across prompts, files, and tool invocations, and validates controls through automated red teaming.

On the repository and infrastructure layers, Rubrik Code Guardian uses frontier AI models from Anthropic to inspect secure duplicates of GitHub and Azure DevOps repositories for complex multi-step attack chains, ranking discovered risks by severity and converting confirmed critical items into tracked engineering tickets. Tanium Cloud delivers real-time visibility across enterprise endpoints to help IT and security teams spot and resolve vulnerabilities, configuration drift, and unapplied system patches. In parallel, Datadog Agent Observability traces agent executions across external tools and backend services, highlighting production usage patterns and recommending optimizations for agent quality, latency, and operational expense.

New agents and products for security: why it matters

Securing autonomous agent workloads requires structural guardrails that traditional perimeter defenses cannot provide on their own. By placing runtime validation, permission policies, and source analysis tools directly inside the Claude ecosystem, organizations can intercept unsafe actions before autonomous operations execute. Cyera Agent Guardian's pre-execution enforcement prevents unauthorized tool calls or data exfiltration across unvetted files, while Chainguard's isolated microVM builds isolate testing environments from broader internal networks during aggressive runtime probe testing.

The consolidation of security verification into developer infrastructure also mitigates operational friction between development speed and security reviews. Netlify Enterprise brings automated preview URLs, role-based access controls, audit logs, and single sign-on to edge agent deployments without requiring manual infrastructure setup. When combined with Rubrik's automated ticket creation for confirmed attack paths and Datadog's execution tracing, engineering organizations obtain end-to-end auditability and immediate remediation paths without forcing developers to switch between disconnected monitoring dashboards.

New agents and products for security: who is affected

Engineering teams, corporate IT departments, and security operations personnel managing production Claude implementations are directly affected by these integrations. Development teams shipping applications via Netlify gain access to managed edge hosting with pre-configured governance controls, while creative teams using Runway, Webflow, and AirOps can generate media assets and manage AI search brand strategies under structured workspace environments. Enterprise users across operations, finance, and project management can deploy prebuilt Asana AI Teammates with individual scoped identities, or use monday agents to control spending, data access, and approvals inside live project boards.

Knowledge workers and cross-functional staff interact with connected enterprise data through integrated conversational interfaces. Zoom Agentic Search allows employees to query meeting summaries, live conversations, documents, and third-party files directly inside Claude without chaining tools or reviewing unformatted call transcripts. Glean links enterprise workflows and repositories across corporate applications, allowing staff to surface answers and execute multi-step work. Enterprise procurement teams also see administrative impact, as qualifying organizations can offset partner software licensing costs against existing Anthropic commercial spend commitments.

New agents and products for security: what to watch

Organizations adopting marketplace integrations must monitor how permission boundaries and identity scopes are defined across autonomous agents. Asana AI Teammates operate under discrete identities and audit trails, routing decisions requiring subjective judgment back to human project owners, which requires ongoing administrative review to prevent unmonitored task bottlenecks. Security administrators will also need to watch red-teaming output from Cyera Agent Guardian to verify whether policy filters on prompts, uploaded files, and tool interfaces hold against adversarial probing.

External developer participation on the marketplace represents another operational area to track. Anthropic maintains an application pathway for outside software vendors building on Claude who wish to list commercial agents and enterprise products within the marketplace catalog. As new infrastructure, security, and productivity services join the existing roster of more than 25 offerings, enterprise administrators must assess how partner data access agreements and endpoint monitoring systems like Tanium Cloud align with corporate compliance requirements.