Show HN: Release-gate – a CLI that checks AI-agent PRs for risky changes

VamsiSudhakaran — Mon, 27 Jul 2026 11:01:39 +0000

I built this after noticing that AI-agent PRs are hard to review. There are many tools available but it does not show the reason why the code can be problematic in case of agent.

A normal diff can show code changes, but it may not make it obvious that a prompt changed, a model changed, a tool config changed, or model output now reaches an eval/exec path.

release-gate is a Python CLI that compares a PR branch against a base ref and reports only what the PR introduced.

Example:

pip install release-gate release-gate pr --base origin/main

It can also create a lockfile for prompts/model/tool config:

release-gate lock . release-gate audit . --lock

This is early. I’m mainly looking for feedback on whether PR-scoped gating is useful, and where the static analysis is still too noisy.

Comments URL: https://news.ycombinator.com/item?id=49067759

Points: 1

# Comments: 0

New comment by VamsiSudhakaran in "Ask HN: What are you working on? (June 2026)"

VamsiSudhakaran — Mon, 15 Jun 2026 14:06:07 +0000

AI agent deployments are happening completely automated nowadays, One command and there it is your agent is live. Recent news about AI agents cost explosions, more than estimated token usage has got me thinking, These are all detected post deployment, pre deployment, are the teams thinking this? Are they even considering who is responsible for what in case of an endless loop being executed that slowly raising the hidden costs.

Introducing - https://release-gate.com/

A pre deployment checker for AI agents that sits in your CI/CD pipeline, No infra required, just a python environment and you are ready to roll.

PS: Comments and suggestions most welcome!!!

Hacker News: VamsiSudhakaran

Show HN: Release-gate – a CLI that checks AI-agent PRs for risky changes

New comment by VamsiSudhakaran in "Ask HN: What are you working on? (June 2026)"