The trust score for AI-generated code. One command verifies what your coding agent shipped: security findings with file:line evidence, Docker-sandboxed build/boot checks, and claim receipts that catch agents lying about tests.
Latest release v1.8.1 — v1.8.1 — the heal/protect release · 5 Aug 2026
1 file for Claude Code, Codex and OpenCode: umbra AGENTS.md — 933 tokens loaded in every session.
AGENTS.md A 933 tok These files are elberacasa/umbra's own configuration — they tell Claude Code, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.