sharathr

5 ms·

1.
by sharathr 2mo ago
we found that using a stable AI-SDLC process like: https://github.com/highflame-ai/ai-factory improves code agent assisted development dramatically specially when going from 1-100.
2.
Show HN: AI-factory – a spec-driven pipeline that stops AI agents rotting code(github.com)
2 points by sharathr 2mo ago|0 comments
3.
by sharathr 2mo ago
part one of the blog is here: https://www.highflame.com/blog/your-ai-gateway-is-fine-until...
4.
How good is your AI Gateway?(highflame.com)
3 points by sharathr 2mo ago|3 comments
5.
by sharathr 2mo ago
We benchmarked Highflame's AI Gateway against popular alternatives to see how our RUST based unified (LLM/MCP) gateway performs and were surprised at how the competition fared (hint: poorly). We wrote a 2 part blog to review the p
6.
Show HN: Deconstructing Anthropic's Coding Agent Control Model(highflame.com)
2 points by sharathr 3mo ago|0 comments
7.
by sharathr 5mo ago
We write about why AI agent sandboxes are insufficient. Identity, not just policy, is critical to securing autonomous AI systems and preventing misuse.
8.
by sharathr 5mo ago
From what I can tell, agent-vault does not solve identity, only how its stored. For true agent identity, you should look into: https://github.com/highflame-ai/zeroid (author: full disclosure)
9.
Autonmomous Agent Identity Framework(github.com)
4 points by sharathr 6mo ago|0 comments
10.
by sharathr 7mo ago
Research paper here: https://arxiv.org/pdf/2602.16935
11.
by sharathr 7mo ago
Hi HN — I’m one of the authors. We’ve been working on security for multi-turn agent loops and noticed most detection approaches operate on isolated prompts. This paper introduces a framework for modeling intent trajectories across sequences
12.
Multi-Turn Intent Detection for LLM and Agent Security (ArXiv)(arxiv.org)
1 points by sharathr 7mo ago|1 comments
13.
by sharathr 9mo ago
We’ve spent decades hardening software supply chains — signing binaries, scanning dependencies, locking down CI/CD — yet AI models themselves are mostly treated as opaque blobs pulled from the internet. That assumption is increasingly
14.
Palisade: Bringing Zero-Trust to the AI Model Supply Chain(highflame.com)
2 points by sharathr 9mo ago|1 comments
15.
by sharathr 1y ago
With zero-friction integration, you can secure code agents at runtime, preventing data leaks, unauthorized actions, and unsafe connections
16.
by sharathr 1y ago
Zero-friction integration to secure code agents at runtime, preventing data leaks, unauthorized actions, and unsafe MCP connections.
17.
by sharathr 1y ago
We recently released https://github.com/getjavelin/ramparts , a tool in the overall ecosystem designed to help protect & raise awareness of security issues in remote or locally developed MCP servers. Its free, exten
18.
by sharathr 1y ago
Looks like Ramparts which solves these issues and is written in fast RUST instead of python. https://github.com/getjavelin/ramparts
19.
by sharathr 1y ago
Palisade works by utilizing dozens of specialized research backed security validators that work together to validate models across different formats (GGUF, SafeTensors, Pickle etc.,) and model families (BERT, Llama etc.,) for things like ba
20.
by sharathr 1y ago
This highlights the critical need for Model Supply Chain scanning for Enterprises that adopt AI. Full disclosure, I am co-founder CEO of Javelin (www.getjavelin.com) and we ran your model through Javelin's Supply Chain Scanner (Palisad
21.
by sharathr 1y ago
yet another observability tool thats joining the already overcrowded space
22.
Ramparts: A fast, lightweight security scanner for mcp servers(github.com)
4 points by sharathr 1y ago|1 comments
23.
by sharathr 1y ago
Ramparts is a fast, lightweight, and robust security scanner specifically engineered for scanning Model Context Protocol (MCP) servers. Its core mission is to identify and mitigate indirect attack vectors and configuration vulnerabilities t
24.
JavelinGuard: Low-Cost Transformer Architectures for LLM Security(arxiv.org)
29 points by sharathr 1y ago|2 comments
25.
You cannot scale, because you can't secure your agents(getjavelin.com)
1 points by sharathr 2y ago|0 comments
26.
by sharathr 3y ago
Not really, prompts are poor guardrails for LLMs and we have seen several examples this fails in practice. We created an LLM focused security product to handle these types of exfils (through prompt/response/url filtering). You can
27.
by sharathr 3y ago
You might want to also look at: https://www.getjavelin.io
28.
by sharathr 3y ago
Have you taken a look at: https://www.getjavelin.io
29.
by sharathr 3y ago
Deploy multiple models, control access to models and give your LLM enabled Apps security guardrails.
30.
Enterprise grade fast LLM Gateway and Router(getjavelin.io)
3 points by sharathr 3y ago|0 comments

More ›