Developer Tools
Choosing canary metrics that can catch a security regression
A canary release limits the damage of a bad change. Here is how to pick metrics, size and duration so it can also surface security regressions.
Oct 18, 20264 min read
Resources
Supply Chain Security, in plain English.
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.
Filter
Articles
RSS feedDeveloper Tools
Can Claude create Excel, Word, PowerPoint and PDF files?
Yes. Claude can create xlsx, pptx, docx and PDF files with code execution turned on. Plans, size limits and what file creation costs in usage.
Oct 18, 20263 min read
Developer Tools
Cache aware input token limits: why caching raises throughput
Some providers do not count cached input tokens against your per minute limit. See which usage fields count, with the arithmetic and one exception.
Oct 18, 20263 min read
Developer Tools
Bring your own LLM key: security for SaaS that accepts customer keys
Letting customers supply their own LLM key moves cost to them but makes you a vault for their secrets. What to store, how, and what to refuse.
Oct 18, 20263 min read
Developer Tools
Bulkheads: isolate tenants and dependencies so one fault stays put
The bulkhead pattern splits resources into pools so one failing dependency or noisy tenant cannot sink everything. How to apply it and where it helps security.
Oct 18, 20263 min read
Developer Tools
Build a searchable code index with SQLite FTS5 in an afternoon
A step by step guide to a local full text search index of your repository using SQLite FTS5 and bm25 ranking, with no servers and no model.
Oct 18, 20263 min read
Developer Tools
Build an internal curated package registry: six steps
How a team can run an internal registry of approved open source packages: admission, builds, updates, removal and what to measure along the way.
Oct 18, 20263 min read
Developer Tools
Build a small eval before you switch to a local model
Thirty real prompts, a pass rule and a side by side run tell you whether a local or quantized model is good enough for your task. A step by step method.
Oct 18, 20263 min read
Developer Tools
How to budget AI coding tools for a team: a pilot method
Do not guess a per developer number. Run a small pilot, use Anthropic's published averages as a scale check, size rate limits and set caps first.
Oct 18, 20263 min read
Stay informed
Weekly insights on software supply chain security, delivered to your inbox.
Self-healing security runs on Safeguard.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.