BLA
BLA (Blind LLM Agents) is the open-source Blue Defense MCP and CLI built to protect software artifacts against automated AI reverse-engineering workflows.
BlueBoot Security runs managed SOCs against simulated attacks and measures how fast and well they detect and respond, so CISOs can compare them head to head.
Translation pending
SecurityHTML · CSSGitHub @nbusichOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Run /nomorevibe verify in your project folder to verify ownership and manage the introduction yourself.
Verify in one minuteAsk about other verification options
BlueBoot Security: head-to-head evaluation of SOCs and blue-team AI agents
Translation pending
Taken from the repository README
Loading the full README.
Read the full README ↗Static site for BlueBoot Security (independent SOC evaluation, plus labeled security datasets for training), built with Bootstrap 5.3 from the jsDelivr CDN and self-hosted Archivo / JetBrains Mono. No build step; served by GitHub Pages from the main branch root.
index.html: the SOC evaluation landing page for CISOs, one page with anchor sections (problem, metrics, how it works, who it's for, contact).ai/index.html: the second landing page, the same method pitched as blue-team evals for AI agents (AI safety, Inspect). Its hero is the agentic SOC A from the demo model, as the homepage hero is SOC B; both are checked by the tests. Paths there start with ../.custom.css: brand tokens (light and dark), type, the hero evaluation replay and section layouts, layered on top of Bootstrap. Blue means detected, orange means missed; keep orange for that only.logo.svg, favicon.svg: the logo and the shield icon.fonts/: Archivo (variable width and weight) and JetBrains Mono, both SIL Open Font License, shared by the site and the demo so the demo renders offline.demo/: the interactive event demo (below). demo/demo.css uses the same tokens and type as custom.css: the navy evaluation console, blue for authorized or detected, orange for attack or missed, dashed outlines for illustrative content.demo/ is an evaluation console: the visitor runs a BlueBoot evaluation as a CISO watching the operation.
Serve the site (python3 -m http.server 8000 --bind 127.0.0.1) and visit /demo/. Opening demo/index.html straight from disk also works, but browsers block the bundled fonts there and fall back to system fonts. It has no dependencies, build step or network requests; fonts come from fonts/. Its CSS and scripts are loaded as demo.css?v=YYYYMMDD and so on (and the homepage's custom.css the same way). GitHub Pages lets browsers cache files for 10 minutes, so bump the version whenever those files change, or a visitor can get the new page with old cached scripts. After 90 seconds idle on the setup or report screen it runs the day again (not under reduced motion). Fullscreen is in the control bar.
demo/core.js holds the whole model: network, attack steps, SOC profiles, per-step status (stepFlow, stepStatus) and scoring. Everything on screen and in the report is computed from one event list it compiles, so they can't disagree. The SOCs, bank and logs are simulated and labeled as such; the profiles are deliberate trade-offs (fastest is not best). The homepage hero shows SOC B from this model, and a test fails if the two drift apart.
node --test demo/tests/sim.test.cjs
node --check demo/app.js
The browser check drives the full flow with all network access blocked and checks five widths for horizontal overflow. Start headless Chrome with --remote-debugging-port=9223 and a temporary profile, then run node demo/tests/browser-check.mjs <screenshot-dir> (Node 22+).