Deep dives, practical guides, and incident analyses from engineers who build Safeguard. No fluff, no vendor FUD — just what you need to ship secure software.
How to read scopes like patient/Observation.rs and system/*.cruds: the three prefixes, the five permission letters, ordering rules and launch scopes.
How to read SMART on FHIR scopes: patient, user and system prefixes, the c r u d s letters, launch context, refresh tokens and what to request.
How server to server FHIR access works with SMART Backend Services: client credentials, a signed JWT, system scopes and short lived tokens, no refresh.
How EHR and standalone SMART launches differ, and the checks both need: PKCE with S256, a high entropy state, the aud parameter and fixed redirect URIs.
Small language models can handle narrow, high-volume tasks at lower cost and latency. How to decide, with examples from published model reports.
What a burn rate is, why multiwindow alerts beat simple error thresholds, and the starting windows Google's SRE workbook suggests for a 99.9% SLO.
How to authorize the Safeguard CLI on a server with no browser: the device code flow, the no browser option, token storage, and what to do about CI.
Two Git config lines let you sign commits with an existing SSH key, needing Git 2.34 or later. Setup, default signing and what a signature proves.
Vendoring copies dependency source into your repo, which gives reviewable, offline builds but adds update and scanning work. Use this guide to decide.
Weekly insights on software supply chain security, delivered to your inbox.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.
