SAN FRANCISCO, July 28, 2026 (GLOBE NEWSWIRE) -- Cycode, the leader in Agentic Development Security, today announced Agentic Workflows, a new capability that puts AI agents to work detecting, prioritizing, and fixing risks the instant they appear, inside boundaries security teams control. Agentic Workflows advance the operating model from human-driven and agent-assisted to agent-driven and human-controlled.
Development moves at machine speed. Now security does too.
AI creates risk faster than any human-driven security process can manage it. More code, new attack surfaces, new classes of AI-specific vulnerabilities, and a shrinking window between disclosure and exploit have made legacy operating models for security unsustainable.
Cycode Agentic Workflows change the operating model. A security team defines a workflow once: the events that trigger it, the sequence of agent actions it runs, and the confidence thresholds and controls that govern each step. From then on, workflows run the instant triggering events occur, whether that is a new critical CVE in a crown-jewel application, a missed SLA on an exploitable finding, or a high-risk violation a developer has ignored. Security teams determine which actions proceed autonomously and which require human review.
"Risk now moves at machine speed while security stays bottlenecked at human speed, and no team can hire its way out of that gap. Waiting for a person to notice each risk and start each response is no longer a viable way to operate," said Lior Levy, co-founder and CEO of Cycode. "Agentic Workflows remove that constraint without removing control. Agents triage and remediate the moment risk appears, and security teams set the scope and the boundaries of autonomy. This is how security finally becomes as agentic as the development it protects: agent-driven and human-controlled."
From autonomous vulnerability response to backlog burndowns, agents carry the load
To fast-track autonomous risk management, Cycode offers a library of pre-built templates that security teams can use as-is or tune to the program’s own scope and control requirements. Templates cover use cases for autonomous triaging and remediation, burning down backlogs, SLA escalation, exception management, and container remediation.
For example, the exception management template triggers when developers ignore high-risk CVEs with known exploits. The change in status from “open” to “ignored” triggers exploitability analysis. If the outcome of the analysis is no exploitability, the agent recommends accepting the developer exception. However, if the analysis confirms exploitability with high confidence, the workflow can reopen the vulnerability, depending on how the team has scoped the workflow.
This is just one example of the new operating model: security that runs at machine speed the moment a risky event occurs, scoped and controlled by human experts. This is how security scales with agentic development.
See Agentic Workflows at Black Hat USA
Agentic Workflows are in early access. To see them in action at Black Hat USA 2026 in Las Vegas, schedule time with the team at https://cycode.com/event/black-hat-usa-2026/.
About Cycode
Cycode is the leader in Agentic Development Security that secures AI development from prompt to runtime. Through unifying control, context, and autonomy in a single platform, Cycode continuously identifies risk across the AI development lifecycle, governs the AI tools developers use, correlates context across the entire software factory, and deploys and manages agents to prevent risk at AI speed.
Media Contact Emily Gallagher PRSENSE for Cycode emily@prsense.global