System Release Notes

Changelog

Weekly engineering release notes, architectural enhancements, and system updates for the 01Security (Z1Sec) platform.

Latest Release: v1.3.0 (August 28, 2026)
v1.3.0 NEW v1.2.9 v1.2.8 v1.2.7 v1.2.6 v1.2.5 v1.2.4 v1.2.3
v1.3.0 Z1Sec

Release v1.3.0

Latest

This release enhances Continuous Operations with improved execution health, failure handling, and lifecycle management. Continuous workloads now provide better visibility into worker health, execution failures, and persistent operation state.

Highlights

Added heartbeat-based worker health monitoring.
Added automatic failure classification and retryability detection.
Added persistent execution and telemetry tracking.
Added circuit breaker support for dependency failures and rate limits.
Added execution admission and task prioritization capabilities.
Improved Pause, Resume, and Stop lifecycle behavior.

Key Changes (v1.3.0)

  • Added heartbeat and stale worker detection.
  • Added structured failure categories and retryability status.
  • Added enriched continuous execution telemetry.
  • Added circuit breaker, admission control, and priority management.
  • Improved continuous execution lifecycle and recovery handling.
#Continuous Operations#Worker Health#Circuit Breaker#Admission Control#Telemetry
v1.2.9 Z1Sec

Release v1.2.9

This release introduces Continuous Ops Agent integration into the Z1Sec execution framework. Continuous monitoring requests are now handled through a dedicated agent and single worker pod, with lifecycle controls, persistent execution tracking, and real-time frontend updates.

Highlights

Integrated Continuous Ops Agent with the Planner.
Added continuous execution mode with single-pod execution.
Added Pause, Resume, and Stop lifecycle controls.
Added real-time SSE streaming for continuous execution updates.
Added a dedicated Continuous Ops UI for monitoring active runs.
Improved runtime handling and validated the flow through automated and Kubernetes testing.

Key Changes (v1.2.9)

  • Added Continuous Ops Agent to the Planner agent catalog.
  • Added continuous intent detection and execution-mode validation.
  • Added single-node/single-pod provisioning for continuous workloads.
  • Added continuous execution and tick tracking.
  • Added REST APIs for continuous session lifecycle management.
  • Added structured SSE events for live execution updates.
  • Added frontend support for continuous monitoring and controls.
  • Improved worker runtime configuration for long-running operations.
  • Verified Continuous Ops through automated tests and live Kubernetes runtime execution.
#Continuous Ops Agent#Lifecycle Controls#SSE Streaming#Continuous UI#Kubernetes Testing
v1.2.8 Z1Sec

Release v1.2.8

v1.2.8 focuses on integrating and validating the native CAI FastAPI runtime within the Z1Sec execution framework. Worker execution was transitioned toward API-based communication using REST and SSE, with improved runtime readiness, structured execution events, task cancellation, and failure handling. The existing orchestration and Kubernetes-based execution model remains intact while the communication layer with CAI workers has been modernized.

Highlights

Integrated the native CAI FastAPI runtime for API-based worker execution.
Introduced a dedicated runtime client for REST and SSE communication.
Added structured real-time execution and telemetry handling.
Improved worker readiness, cancellation, and failure recovery.
Validated the new execution flow through automated and Kubernetes-based testing.

Key Changes (v1.2.8)

  • Added support for running CAI workers through the native FastAPI runtime instead of relying solely on CLI-based execution.
  • Introduced a dedicated CAI Runtime Client to manage sessions, execution, streaming, and cancellation.
  • Added worker health and readiness validation before task dispatch.
  • Integrated SSE-based execution events for structured progress and telemetry updates.
  • Updated failure handling to recognize runtime and network-level worker failures and trigger existing retry mechanisms.
  • Verified the updated execution architecture through unit tests and live Kubernetes runtime testing.
#FastAPI Runtime#REST & SSE#Worker Readiness#Kubernetes Validation#Telemetry
v1.2.7 Z1Sec

Release v1.2.7

v1.2.7 focuses on evaluating and aligning CAI Manager with the latest CAI 1.1.5 runtime capabilities. A comprehensive architectural review was conducted to identify overlapping functionality, define clear responsibility boundaries between the Planner and the CAI runtime, and establish a long-term validation strategy for future runtime compatibility. The release also introduces a structured conformance testing framework to verify runtime behavior across the complete execution pipeline.

Highlights

Conducted a comprehensive architectural audit of CAI 1.1.5 runtime capabilities.
Defined a refined Planner architecture with clear separation between Specialist and Meta Agents.
Introduced a contract-driven verification strategy for validating Planner, Runtime, Kubernetes, and end-to-end execution.
Designed a multi-level conformance testing framework for long-term runtime compatibility validation.
Established an integration roadmap for adopting native CAI 1.1.5 runtime features while preserving platform-specific orchestration.

Key Changes (v1.2.7)

  • Performed a feature-by-feature comparison between the CAI 1.1.5 runtime and the existing CAI Manager architecture to identify overlapping functionality and integration opportunities.
  • Defined a two-tier agent architecture separating Specialist Agents from Meta Agents, enabling clearer responsibility boundaries for future planner enhancements.
  • Refined the planner execution model by introducing capability-driven execution policies for direct execution, delegated routing, and multi-approach orchestration.
  • Designed a six-level architectural verification pyramid covering runtime behavior, agent resolution, planner contracts, artifact exchange, Kubernetes deployment, and end-to-end pipeline validation.
  • Established runtime behavioral validation scenarios and compatibility testing plans for native CAI 1.1.5 features, including Selection Agent, Orchestration Agent, Session Snapshot, and FastAPI runtime support.
  • Documented the architectural migration strategy and compatibility roadmap for integrating new CAI runtime capabilities while maintaining existing distributed orchestration, DAG execution, and Kubernetes-based workflow management.
#CAI 1.1.5#Architectural Review#Meta Agents#Conformance Pyramid#Compatibility Roadmap
v1.2.6 Z1Sec

Release v1.2.6

v1.2.6 focuses on improving the execution framework through architectural refactoring and reliability enhancements. The execution engine has been modularized to improve maintainability, while new scheduling and recovery capabilities lay the foundation for more resilient parallel task execution.

Highlights

Refactored the execution framework into a modular and maintainable architecture.
Enhanced execution scheduling with improved task management capabilities.
Introduced resilient execution mechanisms for improved fault handling and recovery.
Expanded automated test coverage for execution engine components.
Improved code organization to simplify future development and maintenance.

Key Changes (v1.2.6)

  • Modularized core execution and orchestration components to improve code structure and maintainability.
  • Enhanced task execution lifecycle management with improved scheduling and state handling.
  • Added intelligent retry and recovery mechanisms for more reliable execution workflows.
  • Improved execution resource management and runtime coordination for parallel task processing.
  • Strengthened runtime monitoring and execution observability to improve operational visibility.
  • Refactored supporting configuration and execution modules to simplify future enhancements and reduce technical debt.
#Modular Architecture#Resilient Execution#Task Scheduling#Observability#Refactoring
v1.2.5 Z1Sec

Release v1.2.5

v1.2.5 introduces a resilient distributed execution engine for Parallel CIDR Scanning, featuring configurable worker concurrency, queue-based subnet scheduling, persistent execution with checkpoint recovery, session-isolated storage, and enhanced execution observability. This release improves fault tolerance, execution scalability, and real-time visibility while ensuring efficient resource utilization across Kubernetes worker clusters.

Highlights

Introduced configurable concurrent worker execution for Parallel CIDR Scans.
Added dynamic queue-based subnet scheduling with intelligent worker allocation.
Implemented persistent execution with checkpoint-based recovery across pod restarts.
Added session-isolated execution storage with automatic PVC cleanup after successful execution.
Enhanced execution observability with real-time subnet status tracking, scheduler metrics, and live worker progress.

Key Changes (v1.2.5)

  • Added configurable execution controls for concurrent workers and retry handling.
  • Implemented asynchronous queue-based task scheduling for dynamic subnet distribution.
  • Introduced persistent checkpoint storage using Kubernetes Persistent Volumes.
  • Added automatic execution resumption by skipping previously completed subnet checkpoints.
  • Implemented session-specific execution directories for isolated logs and scan artifacts.
  • Added automatic cleanup of execution storage after successful webhook delivery while preserving checkpoints on failures.
#Distributed Engine#Checkpoint Recovery#PVC Storage#Queue Scheduling#Fault Tolerance
v1.2.4 Z1Sec

Release v1.2.4

v1.2.4 introduces Distributed CIDR Scan Orchestration, automatically partitioning large network ranges into isolated subnet workloads for parallel execution. This release improves scan performance with dynamic worker scheduling, real-time subnet progress tracking, resilient retry handling, and unified result aggregation for faster, scalable network security assessments.

Highlights

Introduced distributed CIDR scan orchestration with automatic subnet decomposition and parallel worker execution.
Added dynamic queue-based scheduling with isolated subnet execution and configurable concurrency.
Enhanced execution visibility through real-time per-subnet progress tracking and scheduler metrics.
Improved reliability with independent retry handling and consolidated reporting across all subnet scans.

Key Changes (v1.2.4)

  • Implemented automatic CIDR partitioning to split large network ranges into executable subnet workloads.
  • Added a centralized asynchronous task queue with configurable worker concurrency for dynamic workload scheduling.
  • Extended the execution pipeline to support subnet-level lifecycle management, status tracking, and execution metrics.
  • Introduced result aggregation logic to merge findings from parallel subnet executions while eliminating duplicate results.
#CIDR Partitioning#Subnet Decomposition#Asynchronous Queue#Result Aggregation
v1.2.3 Z1Sec

Release v1.2.3

v1.2.3 introduces Parallel Multi-CIDR Scanning, enabling Z1Sec to efficiently execute security assessments across multiple network ranges within a single Kubernetes execution pod. This release enhances the planning and execution pipeline with intelligent multi-target detection, reusable prompt templating, bounded parallel execution, and consolidated result aggregation, significantly improving scalability while minimizing Kubernetes scheduling overhead.

Highlights

Introduced Parallel Multi-CIDR Scanning with bounded worker pool execution inside a single Kubernetes pod.
Added intelligent multi-target detection and execution-aware planning using reusable prompt templates.
Implemented dynamic target substitution to decouple planner-generated prompts from execution-time target assignment.
Added consolidated result aggregation for vulnerabilities, CVEs, open ports, and execution metrics across parallel workers.

Key Changes (v1.2.3)

  • Added parallel Multi-CIDR scan execution using an intra-pod bounded worker pool.
  • Introduced automatic CIDR detection, validation, subnet splitting, and target normalization.
  • Implemented template-based prompt generation with dynamic {target} substitution during execution.
  • Added representative fingerprint-aware planning to optimize LLM context while avoiding redundant fingerprinting.
  • Implemented parallel result aggregation with consolidated vulnerability, CVE, and port reporting.
  • Extended pipeline models and execution flow to persist and execute cidr_scan_config across the planning lifecycle.
#Multi-CIDR#Worker Pool#Prompt Templates#Fingerprint-Aware#Vulnerability Aggregation