An open-source runtime safety layer for teams that check AI text and images against their rules
Kastra
New· 4Checks what AI coding agents like Claude or Cursor may do before they act, for engineering teams
About Kastra
Kastra places a runtime authorization layer ahead of AI coding agents like Claude, Cursor and Codex, a checkpoint that every planned action has to clear. Its function is to decide in advance what an agent may do.
Authorization before the action
Instead of reviewing what an agent did after the fact, Kastra compares what the agent intends to do with the rules, ahead of execution. Nothing proceeds until the check has been made.
Limits on code and infrastructure
The guardrails cover two areas: how far agents may alter the code they are working in, and how much they may do to the infrastructure within their reach. The intended users are engineering teams that let AI coding agents loose in their own repositories and systems and want firm limits placed on them.
TrustList summary
Kastra is a runtime authorization layer that sits in front of AI coding agents such as Claude, Cursor and Codex. It decides what an agent is permitted to do before the agent acts. It is for engineering teams that want firm limits on what an agent can change in their codebase and infrastructure.
What Kastra offers
- Pre-action authorization: checks an agent's intended actions against rules before they run.
- Codebase guardrails: limits what agents can change in the code they work on.
- Infrastructure guardrails: limits what agents can do to the infrastructure they can reach.
Who it serves
Engineering teams that run AI coding agents on their own codebase and infrastructure.
Summary written by TrustList from the business's own published information, Nov 2026.
Kastra launches
Trust Score
An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.
Updated 9/21/2026
Request a demo or quote from Kastra
Reviews
Questions & answers
No questions yet. Be the first to ask about Kastra.
Kastra alternatives
Similar listings buyers compare against this one.
A control layer that limits which tools and APIs AI agents can call, for security teams
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- AI ModelsBenchmark scores and community ratings for every major model.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.