⚠ Not built. SG/Sentinel is a published design from May 2026 — “this is how I would build it” — not a product. No plans to build it unless somebody funds it. Read the note →
sg-sentinel.sgit.ai / documents / delegation

SG/Sentinel Delegation And Choke-Points: Designing Code To Run Behind The Gate

TypeArch brief Versionv0.27.58 Date18 May 2026 AuthorDinis Cruz (project lead) and collaborators LicenceCC BY 4.0 Sourceraw markdown · view on GitHub

Summary

The deeper continuation of the codebase-extension brief: concentrate checks at the choke-points between trust zones instead of validating everywhere defensively; design application code knowing it runs behind Sentinel so it can delegate validation and authorisation to it; and treat any behavioural difference between QA and production as a major bug (deploy parity). The "no 404s at the API layer" signal turns the layered architecture into a self-checking system: an error deep in the stack that the gate should have caught is information about where the security model has a gap.

Key concepts

Key ideas

On this site

Deploy parity became the parity matrix; the L1 engine being byte-identical across targets is this brief's same-code principle, moved to JS by the CloudFront constraint.

Read the document

📄 Original document · v0.27.58 · 18 May 2026 · rendered from the raw markdown (the source of truth)