News for hackers. Stuff that matters.

veeduzyl

Re: on Sunday April 26, 2026 @12:09AM (#47907253)
We’re building MindForge Guard — a runtime governance layer for AI-assisted development. The core idea: as AI moves from generation to execution, governance needs to happen at runtime, not only after the fact. The upgraded version is now live, focused on single-agent verifiable governance and practical AI-assisted development workflows. Welcome to try it out — your feedback and suggestions would be highly appreciated.
Re: on Thursday February 26, 2026 @03:25PM (#47171553)
We’ve published the first installable GitHub Release of MindForge Guard (v1.0.1).

This release is distributed directly via GitHub (no npm registry dependency).

What’s included:

Policy init schema alignment

Legacy v=1 auto-migration

Kernel bundled inside Guard

Installable .tgz

One-step installer script

Install:

npm install -g ./veeduzyl-mindforge-guard-1.0.1.tgz

Verify:

guard --version guard init guard validate-policy guard status guard drift status

This version focuses on stable distribution and deterministic install flow.

Release page: https://github.com/veeduzyl-hue/mindforge-guard/releases/tag...

Feedback welcome.

Show HN: Guard – An open-core governance layer for AI-generated code

https://mindforge.run/#how-it-works

Hi HN,Over the past year, we've all gotten used to AI writing code.What we haven't solved is this:Who decides whether that code should ship?Guard is an open-core governance layer that sits between AI agents and execution.It does not generate code. It scores risk, detects drift, enforces policy, and can block execution via exit codes.Core concepts:- Decision Snapshot (what AI intends to...

( Read More... | no comments | 2 points | on HN )

Show HN: Verifiable server roundtrip demo for a decision interruption system

https://github.com/veeduzyl-hue/decision-assistant-roundtrip-demo

This is a read-only demonstration repo showing a full server-side decision interruption roundtrip:REQUIRE_CONFIRM → EXECUTE → REJECTThe demo is intentionally minimal and static. It exists only to make one thing verifiable:how a high-risk decision is interrupted,how a receipt + plan_hash are issued,and how execution is either explicitly confirmed or rejected.The repository includes:a runnable...

( Read More... | no comments | on HN )
Re:Ask HN: When tests keep passing but design stops moving on Tuesday January 13, 2026 @09:47AM (#46601660)
I’m experimenting with a cold guardrail for refactor risk. If it blocks you (or fails to), feedback is here: GitHub Discussion: “Feedback: tell me when it interrupts you”
I’m experimenting with a cold guardrail for refactor risk. If it blocks you (or fails to), feedback is here: GitHub Discussion: “Feedback: tell me when it interrupts you”
Re:Refactoring – Not on the backlog (2014) on Tuesday January 13, 2026 @09:40AM (#46601562)
If you try it and it interrupts you, I’m actively collecting feedback here:

GitHub Discussion: “Feedback: tell me when it interrupts you”

There’s no analytics or tracking. Just paste the output of a local telemetry_report (anonymous, local-only).

I’m mostly interested in when people override the guardrail and why.

Re:Show HN: Bind.ly – Persistent memory for AI across tools on Tuesday January 13, 2026 @09:35AM (#46601475)
This tool intentionally has no usage analytics. Feedback happens via a pinned GitHub Discussion instead: paste a local telemetry summary, nothing else.