Developer builds self-healing framework that embeds app knowledge at build time

A developer has built a tool called the Cognitive Autonomic Framework (CAF) to address a long-standing gap in software reliability: critical architectural knowledge encoded in source code is discarded at deploy time, forcing engineers to reconstruct it manually during outages. The framework uses a compiler to parse the codebase at build time and generate a Runtime Semantic Topology (RST), a compact graph capturing service dependencies, retry policies, failure domains, and permitted repair actions. This lightweight graph is bundled directly inside the container, so the application can reason about its own structure when failures occur rather than relying on external inference. The author argues that current approaches — autoscalers blind to application logic and LLM-based observability tools that expensively hallucinate architecture — fail to address a class of failures rooted in application-level semantics. By embedding self-knowledge at the source, the framework aims to enable faster, more accurate autonomous diagnosis without centralised models or token costs.