Observed arrival · 2026-11-03
Shadow turns sensitive prompts into reversible fakes
Shadow is an open-source tool that replaces PII and secrets with realistic, reversible stand-ins before prompts reach LLMs, then restores them.
- For
- Teams using LLMs with sensitive workplace data
- Worth noticing
- The documentation separates a free Core from a per-monitored-user Hub and explains what the Hub receives, stores, and never sees.
Field notes
The documentation maps several deployment paths: proxy, gateway, Rust embedding, Docker/Kubernetes, and integration with an arbitrary endpoint. It also has dedicated material for file masking, tool calls and agents, intent-aware policy, vault behavior, and the security model. The team-facing Hub is presented separately, with a waitlist and per-monitored-user pricing; the page links comparisons against named alternatives including Presidio, GLiNER, and LLM Guard.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue