LIVE
Mistral Makes Large 4 Available as Open Weights29/10/26 · Mistral AI|Gemini 4 Carbon tops the Artificial Analysis Intelligence Index v529/10/26 · Google|Mistral Large 4 now available on Hugging Face29/10/26 · Mistral AI|Google Introduces "Gemini 4 Carbon" as Next Step in Its Model Lineup28/10/26 · Google|DeepMind unveils Genie 4, a new generation of world model27/10/26 · Google DeepMind|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|New benchmark checks whether LLMs actually test their temporal medical claims29/10/26|An ODE-based sampling method speeds up discrete diffusion models29/10/26|OpenAI revises its Model Spec with five guiding values for AI behavior29/10/26 · OpenAI|Inside Proton's internal clash over its AI pivot29/10/26 · Proton|Mistral Makes Large 4 Available as Open Weights29/10/26 · Mistral AI|Gemini 4 Carbon tops the Artificial Analysis Intelligence Index v529/10/26 · Google|Mistral Large 4 now available on Hugging Face29/10/26 · Mistral AI|Google Introduces "Gemini 4 Carbon" as Next Step in Its Model Lineup28/10/26 · Google|DeepMind unveils Genie 4, a new generation of world model27/10/26 · Google DeepMind|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|New benchmark checks whether LLMs actually test their temporal medical claims29/10/26|An ODE-based sampling method speeds up discrete diffusion models29/10/26|OpenAI revises its Model Spec with five guiding values for AI behavior29/10/26 · OpenAI|Inside Proton's internal clash over its AI pivot29/10/26 · Proton|
Toolofficial publisher

Prompt Security

Tool : A precise action the AI can call · requires prompt-security

Managed guardrail for applications and agents.

Solid for enterprise use, paid.

Docs
Overall score
7.6 /10
Security: nothing to flag
Utility
7.0
Reliability
6.0
Security
No caveat on record
Ease
5 min · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

Security teams struggle to monitor what employees send to AI models and what those models return: sensitive data, company secrets, or prompt injections can flow unchecked. Prompt Security sits between users or applications and AI services to inspect every exchange in real time.

The platform acts as a centralized layer, known as an AI gateway, that analyzes traffic to models without requiring each application to be instrumented separately. It detects sensitive data, secrets, prompt injection attempts, and enforces customizable rules called guardrails.

Prompt Security is designed for security teams in organizations of 500 to 1,000 employees or more that deploy AI tools in production or autonomous agents. The platform also publishes complementary open-source tools: ps-fuzz to test system prompt resilience, ClawSec to secure AI agent skills, and OneClaw to track OpenClaw deployments.

This service is not suited for individuals or small organizations without a dedicated security team, as it requires API integration and the configuration of usage policies tailored to the organization.

What it offers

  • ·Real-time inspection of every prompt and response to detect sensitive data and secrets (paid).
  • ·Detection of prompt injection attempts in AI applications and agents (paid).
  • ·Organization-level customizable guardrails, including a topics detector and customer-defined rules (paid).
  • ·Continuous monitoring dashboard tracking prompt volume and triggered alert categories (paid).
  • ·Open-source tool ps-fuzz to test system prompt resilience against 16 attack types across 16 providers.
  • ·Open-source tool ClawSec to audit and verify the integrity of AI agent skills.
  • ·OneClaw tool to track OpenClaw deployments and analyze agent activity across the organization.
  • ·Weekly educational resources on AI security, including a glossary, blog, and podcast.

In practice

Price
Paid
Commercial use
Commercial use: yes
Available on
api
Host
prompt-security
FR interface
No

Security and data

Rating
Security: nothing to flag
Capabilities
readswritesexecutesnetwork
Your data
Data with the vendor
Maintenance
Active

Links and repository checked on 25 October 2026 · added on 1 June 2023

How to install Prompt Security

1. Request a demo at prompt.security to obtain access to the platform. 2. Integrate the platform via API between your applications or users and the AI services you use. 3. Define your guardrails according to your AI Acceptable Use Policy. 4. Enable continuous monitoring and review the dashboard for the first triggered alerts. 5. Test your system prompts with the open-source tool ps-fuzz (available on GitHub) before going live. 6. Precaution: the platform analyzes prompts in read mode; verify with your legal team which data may transit through a third-party service.
proxy ou API

Official documentation ↗

In the same category

Whole category Monitor →