Protocol Debt: When an External Input Decides What Your System Executes
Year: 2026
Type: preprint
Abstract: Log4Shell, Spring4Shell, Text4Shell and React2Shell share no single technique, but they raise the same review question: what can an external input reach? This preprint proposes "protocol debt", an analytical extension of Cunningham's technical debt, defined as the accumulated gap between the capabilities a component exposes by default and those its consumers believe they have enabled. Patching remains necessary, but a CVE identifies a specific vulnerability; it does not record the trust assumption that joined the layers, nor whether another path preserves it. The paper outlines a verifiable registry of trust assumptions per critical service (accepted input, originator, interpretation, invoked authority, and a test with evidence and review date), a six-step method for teams, and a retrospective example applied to Log4j 2. The thesis is tested against two 2026 cases, PHP2Shell and Java4Shell. Both are memory-safety flaws within a single layer, which the registry does not explain; the paper argues that protocol debt governs blast radius and speed of response, not the existence of each bug. This is an architectural hypothesis with a defined test, not a demonstrated single cause. "Protocol debt" is not a standardized term or a validated metric. The four cases illustrate the thesis and do not prove its prevalence.
Author: Downey, Juan
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7235946698
