Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: A MitM proxy to see what your LLM tools are sending

github.com

Post reply on HN
Print view View original

121–128 of 128 posts

Re: Show HN: A MitM proxy to see what your LLM tools are sending

#121

Post by catlifeonmars »

benreesman wrote: post #117
catlifeonmars wrote: post #54

This tool looks like it unconditionally disables tls verification for upstream requests. It shells out to mitmproxy with "--set", "ssl_insecure=true" This took all of 5 minutes to find reading through main.py on my phone. https://github.com/jmuncor/sherlock/blob/fb76605fabbda351828... Edit: In case it’s not clear, you should not use this.

The thing you want has a kind of academic jargon name (coeffects algebra with graded/indexed monads for discharge) but is very intuitive, and it can do useful and complete attestation without compromising anyone credentials (in the limit case because everyone chooses what proxy to run). https://imgur.com/a/Ztyw5x5

Sorry but you lost me. How are coeffects different from effects? I think I’m missing some steps between monads and credentials. Maybe fill in the blanks?

Re: Show HN: A MitM proxy to see what your LLM tools are sending

#122

Post by catlifeonmars »

monkaiju wrote: post #104
catlifeonmars wrote: post #54

This tool looks like it unconditionally disables tls verification for upstream requests. It shells out to mitmproxy with "--set", "ssl_insecure=true" This took all of 5 minutes to find reading through main.py on my phone. https://github.com/jmuncor/sherlock/blob/fb76605fabbda351828... Edit: In case it’s not clear, you should not use this.

And it's already surpassed my most starred project when it was on GitHub, all the more validating to have moved it to forgejo. If vibecoded stuff with unbelievable security vulns can get so much praise the whole star system doesn't work as a quality filter. Similarly a well crafted README used to help reflect quality, no longer...

I don’t use stars to select dependencies FWIW. I look for age, CVEs and what other reputable projects depend on a repo. Also try to look for other signals, like if claims in the readme don’t match the implementation, or if there’s poor hygiene in the CI workflows. (And yes, I have gotten burned by an otherwise well meaning project with a supply chain vuln). As the saying goes “a little copying is better than a little dependency” (see: https://www.youtube.com/watch?v=PAAkCSZUG1c&t=9m28s).

Re: Show HN: A MitM proxy to see what your LLM tools are sending

#123

Post by konform »

syntaxing wrote: post #63

It’s actually really easy to use mitmproxy as a…proxy. You set it up as a SOCKS proxy (or whatever) and point your network or browser to the proxy. I did this recently when a python tool was too aggressive on crawling the web and the server would reject me. Forced my session to limit 5 requests per second and it worked rather than finding the exact file to change in the library. Just do the same to your browser and t…

Actually intercepting the cleartext data is less trivial (not inherently - the browsers just make it more obscure than it could be) but it can be done.

https://news.ycombinator.com/item?id=46820977

Re: Show HN: A MitM proxy to see what your LLM tools are sending

#124

Post by konform »

ctippett wrote: post #31

As someone who just set up mitmproxy to do something very similar, I wish this would've been a plugin/add-on instead of a standalone thing. I know and trust mitmproxy. I'm warier and less likely to use a new, unknown tool that has such broad security/privacy implications. Especially these days with so many vibe-coded projects being released (no idea if that's the case here, but it's a concern I have nonetheless).

[dead]

Re: Show HN: A MitM proxy to see what your LLM tools are sending

#125

Post by spl757 »

I usually have small mini-pc with at least two ethernet ports and configure it as a transparent bridge sitting between my desktop and the router/switch. Give the bridge a local IP, set up some packet inspection stuff, and you can easily monitor anything and everything going in and out. It's not all I use, but it's one part.

I also run ai models locally and like to verify that things aren't talking to the internet if they aren't supposed to be.

Re: Show HN: A MitM proxy to see what your LLM tools are sending

#128

Post by mgabbardo »

I built something similar after seeing this post: https://wiretaps.ai (repo: https://github.com/marcosgabbardo/wiretaps)

Different approach:

- No TLS verification bypass — works by setting OPENAI_BASE_URL

- Built-in PII detection (SSN, credit cards, emails, phone numbers across ~20 countries)

- Crypto detection (BTC/ETH addresses, private keys, seed phrases)

- SQLite by default, zero config: pip install wiretaps && wiretaps start

Still early (v0.3), but the PII detection is solid — 45+ regex patterns for global compliance (GDPR, LGPD, etc).

Would love feedback from folks here.

Post reply on HN