HackerReposA FIELD GUIDE TO OPEN SOURCE
ProSponsor the shelf♧ 0
← Back to top all-time
borenstein
yolo-cage
Python2026–2026
borenstein / yolo-cage · Archived

yolo-cage

AI coding agents that can't exfiltrate secrets or merge their own PRs.

3Distinct people
6Mentions
0Explicit recommenders
109GitHub stars
REPOSITORY AT A GLANCE

What to know before exploring

What it is

AI coding agents that can't exfiltrate secrets or merge their own PRs.

GitHub

Where it is used

No documented product use has been verified yet.

What people are saying

No recommendation phrase was found in the indexed comments.

No classified warning is attached to this page. Read the discussions before treating silence as approval.

Also in these discussions

No other repository is repeatedly linked in the same discussions.

This week on Hacker News

Nov 9–Nov 16 UTC · Direct repository links and separately labeled related project stories. Indexed through 11/16/2026, 2:19:09 PM UTC.

No direct repository links were found in the indexed HN discussions over the past seven days. Older discussions are available below.

Mentions over time

HN mentions by 6-hour interval6 dated mentions

Observed mentions: 2026-01-13 20:25:59 UTC to 2026-01-21 17:46:08 UTC.

01232026-01-13 18:00:00 UTC to 2026-01-14 00:00:00 UTC (exclusive): 3 mentions; 1 submissions, 2 commentsJan 13, 18:002026-01-14 00:00:00 UTC to 2026-01-14 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-14 06:00:00 UTC to 2026-01-14 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-14 12:00:00 UTC to 2026-01-14 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-14 18:00:00 UTC to 2026-01-15 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-15 00:00:00 UTC to 2026-01-15 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-15 06:00:00 UTC to 2026-01-15 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-15 12:00:00 UTC to 2026-01-15 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-15 18:00:00 UTC to 2026-01-16 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-16 00:00:00 UTC to 2026-01-16 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-16 06:00:00 UTC to 2026-01-16 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-16 12:00:00 UTC to 2026-01-16 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-16 18:00:00 UTC to 2026-01-17 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-17 00:00:00 UTC to 2026-01-17 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-17 06:00:00 UTC to 2026-01-17 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-17 12:00:00 UTC to 2026-01-17 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 commentsJan 17, 12:002026-01-17 18:00:00 UTC to 2026-01-18 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-18 00:00:00 UTC to 2026-01-18 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-18 06:00:00 UTC to 2026-01-18 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-18 12:00:00 UTC to 2026-01-18 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-18 18:00:00 UTC to 2026-01-19 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-19 00:00:00 UTC to 2026-01-19 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-19 06:00:00 UTC to 2026-01-19 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-19 12:00:00 UTC to 2026-01-19 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-19 18:00:00 UTC to 2026-01-20 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-20 00:00:00 UTC to 2026-01-20 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-20 06:00:00 UTC to 2026-01-20 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-20 12:00:00 UTC to 2026-01-20 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-20 18:00:00 UTC to 2026-01-21 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-21 00:00:00 UTC to 2026-01-21 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-21 06:00:00 UTC to 2026-01-21 12:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-01-21 12:00:00 UTC to 2026-01-21 18:00:00 UTC (exclusive): 3 mentions; 1 submissions, 2 commentsJan 21, 12:00

Counts come from the repository’s dated source evidence. Index cutoff: 2026-11-02 08:16:59 UTC.

View mention counts and date ranges
UTC ranges include their start and exclude their end. Select a period to filter the source evidence.
PeriodStart (UTC)End (exclusive, UTC)MentionsSubmissionsComments
Jan 13, 18:002026-01-13 18:00:00 UTC2026-01-14 00:00:00 UTC312
Jan 14, 00:002026-01-14 00:00:00 UTC2026-01-14 06:00:00 UTC000
Jan 14, 06:002026-01-14 06:00:00 UTC2026-01-14 12:00:00 UTC000
Jan 14, 12:002026-01-14 12:00:00 UTC2026-01-14 18:00:00 UTC000
Jan 14, 18:002026-01-14 18:00:00 UTC2026-01-15 00:00:00 UTC000
Jan 15, 00:002026-01-15 00:00:00 UTC2026-01-15 06:00:00 UTC000
Jan 15, 06:002026-01-15 06:00:00 UTC2026-01-15 12:00:00 UTC000
Jan 15, 12:002026-01-15 12:00:00 UTC2026-01-15 18:00:00 UTC000
Jan 15, 18:002026-01-15 18:00:00 UTC2026-01-16 00:00:00 UTC000
Jan 16, 00:002026-01-16 00:00:00 UTC2026-01-16 06:00:00 UTC000
Jan 16, 06:002026-01-16 06:00:00 UTC2026-01-16 12:00:00 UTC000
Jan 16, 12:002026-01-16 12:00:00 UTC2026-01-16 18:00:00 UTC000
Jan 16, 18:002026-01-16 18:00:00 UTC2026-01-17 00:00:00 UTC000
Jan 17, 00:002026-01-17 00:00:00 UTC2026-01-17 06:00:00 UTC000
Jan 17, 06:002026-01-17 06:00:00 UTC2026-01-17 12:00:00 UTC000
Jan 17, 12:002026-01-17 12:00:00 UTC2026-01-17 18:00:00 UTC000
Jan 17, 18:002026-01-17 18:00:00 UTC2026-01-18 00:00:00 UTC000
Jan 18, 00:002026-01-18 00:00:00 UTC2026-01-18 06:00:00 UTC000
Jan 18, 06:002026-01-18 06:00:00 UTC2026-01-18 12:00:00 UTC000
Jan 18, 12:002026-01-18 12:00:00 UTC2026-01-18 18:00:00 UTC000
Jan 18, 18:002026-01-18 18:00:00 UTC2026-01-19 00:00:00 UTC000
Jan 19, 00:002026-01-19 00:00:00 UTC2026-01-19 06:00:00 UTC000
Jan 19, 06:002026-01-19 06:00:00 UTC2026-01-19 12:00:00 UTC000
Jan 19, 12:002026-01-19 12:00:00 UTC2026-01-19 18:00:00 UTC000
Jan 19, 18:002026-01-19 18:00:00 UTC2026-01-20 00:00:00 UTC000
Jan 20, 00:002026-01-20 00:00:00 UTC2026-01-20 06:00:00 UTC000
Jan 20, 06:002026-01-20 06:00:00 UTC2026-01-20 12:00:00 UTC000
Jan 20, 12:002026-01-20 12:00:00 UTC2026-01-20 18:00:00 UTC000
Jan 20, 18:002026-01-20 18:00:00 UTC2026-01-21 00:00:00 UTC000
Jan 21, 00:002026-01-21 00:00:00 UTC2026-01-21 06:00:00 UTC000
Jan 21, 06:002026-01-21 06:00:00 UTC2026-01-21 12:00:00 UTC000
Jan 21, 12:002026-01-21 12:00:00 UTC2026-01-21 18:00:00 UTC312

What Hacker News says

3 sources
It depends what your threat model is and where the container lives. For example, k8s can go a long way towards sandboxing, even though it's not based on VMs. The threat with AI agents exists at a fairly high level of abstraction, and developing with them assumes a baseline level of good intentions. You're protecting against mistakes, confusion, and prompt injection. For that, your threat mitigation strategy should be focused on high-level containment. I've been working on something in a similar vein to yolobox, but the isolation goal has more to do with secret exfiltration and blast radius. I'd love some feedback if you have a chance! https://github.com/borenstein/yolo-cage
In Show HN: Yolobox – Run AI coding agents with full sudo without nuking home dir
This is great - I've been working on a related but different approach called yolo-cage (https://github.com/borenstein/yolo-cage) that addresses a different threat model. Yolobox protects your local machine from accidental damage (crucial if you're worried about rm -rf ~). yolo-cage focuses on secret exfiltration and multi-agent coordination: it runs in Kubernetes, scans all egress traffic for secrets (API keys, tokens, SSH keys), enforces git branch isolation so multiple agents can work in parallel, and ensures agents can't merge their own PRs. "Agent proposes, human disposes." The escape testing framework has a meta twist: there's a canned prompt that asks Claude to try to escape its own containment, and that prompt lives inside the repository. The agent verifies the prompt is legitimate by finding it in the repo it's auditing - the cage
Read the rest
contains its own escape room instructions. (I tried to post this as a separate Show HN but my account is too new - happy to discuss the tradeoffs between local sandboxing vs. server-side containment here.)
In Show HN: Yolobox – Run AI coding agents with full sudo without nuking home dir

MAKE YOURSELF AT HOME

A few shortcuts

/
Search the collection
?
Open this guide
Esc
Close dialogs and navigation

Your saved repositories and interests stay in this browser.