Raviteja_

4 ms·

1.
by Raviteja_ 7mo ago
This is Ravi, the developer who built this. Few months back, I was building an AI app and realized my users data was going to OpenAI in plaintext. I searched for a local firewall something that would sit between my app and the LLM provider
2.
by Raviteja_ 7mo ago
I open sourced Sentinel Protocol today. 52,069 lines of code. 81 security engines. 9 runtime dependencies. MIT. It's a local proxy that sits between your app and any LLM (OpenAI, Anthropic, Gemini, Ollama) and enforces PII protection,
3.
by Raviteja_ 7mo ago
I've been quietly building this forlast few months. Today I am open-sourcing it. One thing that kept bothering me is that teams I worked with were sending raw user input to OpenAI with literally zero filtering. Not even a regex. Users
4.
Sentinel Protocol – open-source AI firewall
1 points by Raviteja_ 7mo ago|0 comments
5.
by Raviteja_ 8mo ago
3 Tools you are overpaying for: - Prompt Compressor ($10/mo) - PII Redactor ($20/mo) - Fake Data Gen ($15/mo) Risk Mirror does all 3 in the browser. The Text Suite is $0 right now. No card required. No logs kept. Try for fre
6.
by Raviteja_ 8mo ago
Every time you paste a stack trace into ChatGPT, you might be leaking: - User session tokens - Database connection strings - API keys from env variables I built Risk Mirror to scan and redact sensitive data BEFORE it touches any AI. It&#x27
7.
by Raviteja_ 8mo ago
Stop burning $20/mo on Claude credits. I unlocked the Prompt Optimizer for free. If you're hitting the message cap on Claude/Cursor, you're sending too much fluff. "Please", "Thank you," and verbose c
8.
by Raviteja_ 8mo ago
If your Cursor/Claude credits vanish fast, it’s probably prompt bloat. Polite filler + repeated context + messy JSON = wasted tokens. Risk Mirror compresses prompts 20–40% without changing meaning. More credits. Same results. Try free:
9.
by Raviteja_ 8mo ago
Hey all, I'm the creator of the PII Firewall Edge API (currently on RapidAPI). I saw a lot of devs struggling to implement safety guardrails correctly—most were just using basic regex or heavy LLMs that hallucinate. So, I decided to pa
10.
Stop leaking user data to OpenAI/Claude/Gemini(risk-mirror.vercel.app)
1 points by Raviteja_ 8mo ago|4 comments
11.
by Raviteja_ 9mo ago
Great catch! Emails with spaces around @ (like "test @ example.com") slip through. This is a classic obfuscation bypass. The current pattern intentionally matches RFC 5321 compliant emails (no spaces). Adding support for spaced va
12.
by Raviteja_ 9mo ago
Quick technical notes for HN: Why no AI? The irony of sending PII to an AI model to detect PII is lost on most "privacy" APIs. This is pure algorithmic detection – the same approach your credit card company uses to validate card n
13.
Show HN: I built a zero-log PII redaction API – no AI, just regex and checksums(pii-firewall-edge-web.vercel.app)
1 points by Raviteja_ 9mo ago|4 comments
14.
by Raviteja_ 9mo ago
Hi HN! I built this after 3 months researching image-based attacks. The problem: Apps that accept user images typically just strip EXIF metadata. But this misses: - Steganographic payloads (data hidden in pixel LSBs) - Polyglot files (vali
15.
Show HN: Zero Trust API – Image CDR in Rust/WASM (Rebuild Images from Pixels)(zero-trust-web.vercel.app)
2 points by Raviteja_ 9mo ago|1 comments