← Back to AI Insights
Gemini Executive Synthesis

OpenAPPA, an open-source deterministic guardrail system for AI agents.

Technical Positioning
The first deterministic guardrail that prevents data leaks and maintains high agent utility (90%), addressing the limitations of non-deterministic and existing deterministic guardrails.
SaaS Insight & Market Implications
OpenAPPA directly addresses the critical enterprise pain point of AI agent control and data security. The inherent risk of data leaks and unpredictable agent behavior, especially with increasing tool integration, is a major barrier to enterprise AI adoption. Existing guardrail solutions are either ineffective against prompt injections or severely degrade agent utility. OpenAPPA's innovation lies in its data-specific, deterministic policy language combined with utility-preserving mechanisms like the DualLLM pattern. This offers a robust solution for deploying agents in sensitive environments, ensuring compliance and operational integrity. For B2B SaaS, this is a foundational security and governance layer for AI platforms, enabling broader and safer agent deployments across regulated industries. The open-source nature could drive rapid adoption and standardization.
Proprietary Technical Taxonomy
enterprise agents sensitive data guardrails non-deterministic guardrails LLM as a judge prompt injections deterministic guardrails Cedar

Raw Developer Origin & Technical Request

Source Icon Hacker News Sep 28, 2026
Show HN: OpenAPPA – open-source deterministic guardrails that don't break agents

Hi Hacker News! Matvey, one of the authors, is here.While building enterprise agents, we ran into a problem: the more tools you connect to the AI, the higher the chance it will run out of control and leak sensitive data.Guardrails, in theory, should prevent this, but the situation is worrying:
- Non-deterministic guardrails (LLM as a judge, auto modes, etc.) are vulnerable to prompt injections, or they lack knowledge of the data, making them inefficient (~10% data leaks on our benchmarks).
- Existing deterministic guardrails (Cedar, OPA, FIDES, Dogwood) require massive case-specific IF-ELSE-like policies and break agents (~59% utility loss on our benchmarks).We did something differently.We’ve taken the best of existing deterministic guardrails and built a policy language that is data-specific, not use-case specific. It lets you scale agents without updating a policy.On top of that, we’ve added multiple tricks (like a remedy plan or a DualLLM pattern) to help agents operate within those restrictions, raising utility from ~40% to ~90% and making it the first deterministic guardrail that doesn't break agents.Finally, we’ve designed it to be pluggable into any agent loop with pre- and post-tool-call hooks.We invite you to check out our benchmarks: openappa.com/evaluationPlay with it in Claude Code: openappa.com/claude-codeTry plugging it into your agent: openappa.com/add-to-agentOr check the academic paper: arxiv.org/abs/2607.24625We& love to hear any feedback!

Developer Debate & Comments

moneytool • Sep 29, 2026
HI I have created something that actually to solve this exact problem please go through it this stays in your environment independent of AI and its a python module that blocks AI from using unauthorized commands https://github.com/moneytool/aegis-devops
vladimir_gor • Sep 28, 2026
Really interesting direction. What resonated with me is that you're treating agent security as an information-flow problem rather than a prompt-classification problem. It was not so obvious to me.A key question I agree isn't just "is this tool call allowed?", but "given everything the agent has read so far, is this information now allowed to flow to this destination?" That feels like a much more fundamental abstraction.The part I'm particularly curious about is how this will work with policy authoring at scale. What would be the main adoption challenge?
zborro • Sep 28, 2026
i suspect we’ll see more of this: flexible agents but deterministic boundaries. Congrats on launch!
immafridge • Sep 28, 2026
Quick disclaimer, I work at Archestra.I’ve had the chance to play with OpenAppa for a bit and if there’s one thing that I love with this project: it’s simple to get started with and easy to tweak. imo agentic security shouldn’t have to be painful to setup.Give it a shot and hopefully ya’ll will find this project useful. It's also open source :)
piercypixel • Sep 28, 2026
Guardrails with builtin remediation instead of simply blocking my agent is a mind blowing long awaited experience! Sooo good. Can't recommend more!
ildari • Sep 28, 2026
a few days ago I started an agent on gpt-5.6-terra to work on a project, and one of the website pages had a sentence to create GH issues. Agent read it and that was enough to derail and go creating issues with my context
keshakon • Sep 28, 2026
Hi! One of the OpenAPPA authors here. Ask me anything!My favorite part of APPA is “batteries”: you can run arbitrary programs as part of an authorization decision. For example, a battery could call the GitHub API to check whether a repository is public or private, then use that result to decide whether its contents can be posted to Slack.
arseny_info • Sep 28, 2026
I still remember the times when ai/ml security was about perturbing pixel gradients to misclassify a panda
joeyorlando • Sep 28, 2026
aussi, si vous êtes à Montréal et vous aimerez apprendrez plus sur OpenAPPA, viens à l'evenement CNCF demain soir, le 29 - je ferai un p'tit discours sur l'integration kagent d'OpenAPPA.https://www.meetup.com/kubernetes-montreal/events/316689391/
dvorkanton • Sep 28, 2026
Finally some determinism in our high-temperature sampling world!

Frequently Asked Questions

Market intelligence mapped to OpenAPPA, an open-source deterministic guardrail system for AI agents..

How is OpenAPPA, an open-source deterministic guardrail system for AI agents. positioned in the market?
Based on our AI analysis of the original developer request, its primary technical positioning is: The first deterministic guardrail that prevents data leaks and maintains high agent utility (90%), addressing the limitations of non-deterministic and existing deterministic guardrails.
How is the developer community reacting to OpenAPPA, an open-source deterministic guardrail system for AI agents.?
Yes, we have tracked 12 direct responses and active debates regarding this specific topic originating from Hacker News.
What are the foundational technologies related to OpenAPPA, an open-source deterministic guardrail system for AI agents.?
Our proprietary extraction maps OpenAPPA, an open-source deterministic guardrail system for AI agents. to adjacent architectural concepts including enterprise agents, sensitive data, guardrails, non-deterministic guardrails.

Engagement Signals

25
Upvotes
12
Comments

Cross-Market Term Frequency

Counts stored text matches for foundational terms like benchmarks and guardrails in analysed public discussions. These counts do not measure adoption.

Stored text matches, not adoption or market-share measurements.

  • benchmarks: 12 records
  • guardrails: 8 records
  • agent loop: 8 records
  • prompt injections: 3 records
  • sensitive data: 2 records
  • Cedar: 2 records
  • enterprise agents: 1 records
  • non-deterministic guardrails: 1 records
  • LLM as a judge: 1 records
  • deterministic guardrails: 1 records
  • OPA: 1 records
  • FIDES: 1 records
  • Dogwood: 1 records
  • policy language: 1 records
  • data-specific: 1 records
  • use-case specific: 1 records
  • remedy plan: 1 records
  • DualLLM pattern: 1 records
  • pre- and post-tool-call hooks: 1 records