Custosa
Permission and compliance layer for Enterprise AI
What it does
Every enterprise AI deployment has the same quiet vulnerability. Your LLM can see everything. Contracts, patient records, classified specs. It doesn't know who's asking or what they're allowed to see. Custosa enforces permissions before the model touches context. Not output filters. Upstream access control, compiled at runtime from user identity and compliance policy. Same query. Different context. Depending on who's asking. HIPAA, ITAR, SOC2 ready. Built for RAG and agentic pipelines.
Does the same job
all alternatives →
AI Hive May 2026 · app.aihive.global · ▲53Build, deploy, and run all your AI agents in one platform.
Custos — The bouncer for your AI agents28d ago · sanjaynandanj.github.io · ▲7Stop AI agents from doing things they shouldn't.
- CLCompliant LLM toolkit for ensuring compliance & security of AI systems2025 · github.com · ▲7
With the right technique, I was able to break the so-called secure models like Claude and OpenAI. So, I built an open-source tool to automate this and find security holes in any hosted model. I got claude-sonnet-4 to demonstrate the following harmful behavior: - steal data from downstream tool calls using sql injection, code injection and template injection attacks - install spyware or malware using prompt obfuscation to send data to a third-party server Try it yourself with this simple command: pip install compliant-llm && compliant-llm dashboard
- CCControl Claude permissions using a cloud-based decision table UIJan 2026 · github.com · ▲14
We’ve been building visual rule engines (clear spreadsheet interfaces -> API endpoints that map incoming data to a large number of potential outcomes), and had the fun idea lately to see what happens when we use our decision table UI with Claude’s PreToolUse hook. The result is a surprisingly useful policy/gating layer– these tables let your team: - Write conditional, exception-friendly policies beyond globs/prefixes (e.g. allow rm -rf only in */node_modules/*, deny / or $HOME, ask if --force or network call; gate kubectl delete / SQL DROP with a clear reason) -…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, March 2026
the whole month →
Switch from ChatGPT to Claude with import memory feature
AI · Mar 2026 · claude.com
