HackerLangs
TopNewTrendsCommentsPastAskShowJobs

syumei

78 karmajoined 4 years ago
Ph.D. student @columbia working on Security, Crypto, and AI/ML

Submissions

A Cursor Sandbox Escape Shows Why AI Agents Need Kernel Boundaries

medium.com
1 points·by syumei·3 days ago·0 comments

A Conflict-Free Multi-Agent Ensemble for Claude and Codex

medium.com
1 points·by syumei·7 days ago·2 comments

CurXecute: When a Slack message rewrites your editor's config and runs code

medium.com
3 points·by syumei·7 days ago·0 comments

Claude Code SOCKS5 Bypass Shows Why Egress Filters Belong at the Boundary

medium.com
1 points·by syumei·8 days ago·0 comments

Reached 440+ stars: Auditable sandbox to record what AI agents did

old.reddit.com
2 points·by syumei·9 days ago·0 comments

Auditable Workspaces for AI Coding Agents

medium.com
1 points·by syumei·10 days ago·0 comments

I built an anime-style UI for watching AI coding agents review each other's code

old.reddit.com
3 points·by syumei·12 days ago·0 comments

How Claude Code and Codex Sandbox Untrusted Code

medium.com
2 points·by syumei·13 days ago·1 comments

Two Heads Are Better Than One: Run Many AI Agents, Merge One Auditable Result

medium.com
2 points·by syumei·18 days ago·0 comments

How Claude Code Broke My Git Worktree

medium.com
3 points·by syumei·24 days ago·0 comments

A Git/LFS-backed unified tool-output format for AI agents, with 95% token cut

old.reddit.com
2 points·by syumei·last month·0 comments

Claude Code and Codex can have real-time conversation via Git

medium.com
116 points·by syumei·last month·79 comments

[untitled]

1 points·by syumei·last month·0 comments

Git-based cache saves 50% on token usage

old.reddit.com
34 points·by syumei·2 months ago·1 comments

[untitled]

1 points·by syumei·3 months ago·0 comments

[untitled]

1 points·by syumei·3 months ago·0 comments

Hacking MCP Servers in AI Systems – The Rug Pull: Tool Changes After Approval

medium.com
2 points·by syumei·3 months ago·0 comments

[untitled]

1 points·by syumei·3 months ago·0 comments

[untitled]

1 points·by syumei·3 months ago·0 comments

[untitled]

1 points·by syumei·3 months ago·0 comments

comments

syumei
·7 days ago·discuss
Thank you so much!
syumei
·13 days ago·discuss
Claude Code and Codex both run untrusted commands. This post compares how their sandboxes protect your repo, secrets, and host machine.
syumei
·24 days ago·discuss
quite interesting post
syumei
·last month·discuss
Fair criticism. I don’t think the value is just "agents can send text to each other”; that part has many implementaton and design choices. The value of h5i I’m exploring is making the intermediate state reviewable: review requests, risks, handoffs, unresolved claims, associated prompts and AI-to-Ai conveersation, and final decisions tied to the branch/PR.
syumei
·last month·discuss
[dead]
syumei
·last month·discuss
[dead]
syumei
·last month·discuss
[flagged]
syumei
·last month·discuss
[dead]
syumei
·last month·discuss
[dead]
syumei
·3 months ago·discuss
[dead]
syumei
·3 months ago·discuss
I've been tracking real-world AI agent failures and incidents recently.

Things like:

prompt injection → goal hijacking

agents going rogue due to misalignment

unintended/unsafe tool use

It feels like we're starting to see repeatable patterns, not just isolated bugs.

I'm collecting cases + papers here:

https://github.com/h5i-dev/awesome-ai-agent-incidents

If you've seen interesting incidents, weird failures, or relevant research, I would love to add them.