Autoptic expands change resilience catalog to 250 components

Autoptic on Aug. 19, 2026, said its Systems Change Resilience Catalog has grown to 250 reusable components, up from 95 in May, as the company pushes preventative AI SRE for engineering teams. The expanded library is aimed at helping teams detect risk earlier, automate operations, and turn customer-specific work into reusable software building blocks.

Why it matters: - The expanded catalog gives engineering, platform, DevOps and SRE teams a larger reusable base for spotting production risk before incidents reach customers. - Autoptic is packaging customer learnings into components that can be reused across organizations, which can shorten time-to-value and reduce repeated custom work. - The catalog is positioned as a foundation for preventative AI SRE, not just incident response.

What happened: - Autoptic announced that its Systems Change Resilience Catalog now includes 250 reusable components. - The catalog breaks down into 21 Data Sources, 110 Tools, 94 Skills and 25 Agents. - The company first unveiled the Change Resilience Catalog in May 2026 with 95 components. - The original catalog included 15 Data Sources, 40 Tools, 30 Skills and 10 Agents. - The announcement was made in Austin, Texas, on Aug. 19, 2026.

The details: - The 21 Data Sources connect Autoptic to telemetry, events, issues, changes, repositories and other engineering data without relocating it. - Integrations include AWS services such as CloudWatch, CloudTrail and Cost Explorer. - The Data Sources also include observability systems such as Datadog and Grafana’s ecosystem, including Prometheus, Loki and Tempo. - Delivery and tracking systems covered include GitHub, GitLab and Jira. - The 110 Tools provide deterministic analysis across those sources. - Each Tool is a versioned Performance Query Language program that returns charts, notes and ranked entities. - The Tool set spans cloud infrastructure such as EC2, Lambda and S3. - The Tool set also reaches Kubernetes internals including etcd, HPA and PVC. - Additional coverage includes PostgreSQL, Kafka, Redis, JVM and the Node.js event loop. - The company says the Tools are authored as ordinary code, making them reviewable, diffable and manageable through AI coding systems like Claude. - The 94 Skills combine Tools into repeatable operational workflows for DevOps and SRE teams. - Examples include triage playbooks for services and databases, Kubernetes diagnostics for autoscaler and PVC-fill issues, and GitHub Actions pipeline health checks. - Skills support fixed workflows with schedules and notification floors, as well as more exploratory Detective pathways that classify operator intent and route to the right Skill. - The 25 Agents combine Skills, Tools, prompts, AI inference and orchestration to deliver higher-level operational functions. - Agent families cover platform and orchestration reliability, database systems, infrastructure and networking, and release management. - Agents power timed and triggered Briefs and on-demand Investigations. - Autoptic says customer deployments and implementations have expanded every layer of the catalog since May.

Between the lines: - The catalog growth suggests Autoptic is using a product flywheel: customer problems become reusable components, and those components then accelerate future deployments. - The structure also signals a modular approach to AI SRE, with each layer handling a different job from data collection to computation to workflow execution to orchestration. - Autoptic is emphasizing control and transparency, likely to distinguish its system from more opaque agentic tools.

What’s next: - Autoptic plans to keep expanding the catalog as new customer patterns are identified and converted into reusable components. - The company says the platform will continue to support both Autoptic-hosted and Bring Your Own Cloud deployments. - Autoptic also plans to maintain support for commercial and open-source models while letting customers control data, infrastructure, permissions, workflows and inference costs.

The bottom line: - Autoptic is turning one-off production engineering work into a larger reusable library, betting that broader component coverage will make preventative AI SRE more practical for enterprise teams.