stack.tools

Guardrails & AI Security

Safety filters, prompt-injection defense, PII redaction, AI security vendors and practices.

12 verified facts across 8 companies

Cursor (Anysphere)

AI code editor and agentic coding platform (Agent, Tab, Composer, Bugbot, Cloud Agents).

Sandboxed terminals (Seatbelt / seccomp + Landlock) by Anysphere

Runs agent shell commands in OS-level sandboxes with restricted filesystem access.

High confidence

Receipts · 2

Show quotes (2)
  • “Seccomp blocks unsafe syscalls, while Landlock enforces filesystem restrictions.” — cursor.com, Feb 18, 2026
  • “A mistaken agent can delete databases, ship broken code, or leak secrets.” — cursor.com, Feb 18, 2026

Source last verified Jul 26, 2026

Klarna

Swedish buy-now-pay-later payments and consumer banking company (NYSE: KLAR).

GenAI governance tooling by Klarna (in-house)

Gates GenAI use for safety and fintech compliance.

High confidence

Receipts · 1

Show quotes (1)
  • “own tooling and processes that enable safe, compliant, and efficient GenAI use” — jobsinforex.com, Jul 21, 2026

Source last verified Jul 26, 2026

AI Assistant topic guardrails by Klarna (in-house)

Restricts the assistant to supported topics with human handoff on sensitive queries.

Medium confidence

Receipts · 1

Show quotes (1)
  • “When an "out of bounds" topic comes up, the chatbot hands over to a human agent” — blog.pragmaticengineer.com, Feb 29, 2024

Source last verified Jul 26, 2026

Lovable

AI app builder turning prompts into full applications.

Modal Sandboxes by Modal

Runs LLM-generated app code in safe, isolated sandboxes.

High confidence

Receipts · 1

Show quotes (1)
  • “Modal Sandboxes are now used to serve every app generation session in Lovable.” — modal.com, Jul 7, 2025

Source last verified Jul 27, 2026

Notion

Connected AI workspace for docs, wikis, projects, and enterprise search (Notion AI, Agents, Q&A).

Zero-retention LLM contracts by Notion

Enforces zero-retention and no-training contracts with AI subprocessors.

High confidence

Receipts · 2

Show quotes (2)
  • “our LLM providers utilize zero data retention for Enterprise plan workspaces” — notion.com, Jul 2026
  • “prohibit the use of Customer Data to train their models” — notion.com, Jul 2026

Source last verified Jul 26, 2026

Vercel Sandbox by Vercel

Runs untrusted Notion Workers code in isolated sandboxes.

High confidence

Receipts · 1

Show quotes (1)
  • “Under the hood, every Worker runs on Vercel Sandbox .” — vercel.com, Mar 12, 2026

Source last verified Jul 27, 2026

Runlayer by Anysource (Runlayer)

Secures Notion's MCP connections and agent tool traffic.

Medium confidence

Receipts · 1

Show quotes (1)
  • “Security for MCP connections” — registora.com, Jul 9, 2026

Source last verified Jul 26, 2026

Perplexity

AI answer engine combining a proprietary web index with LLMs to deliver cited, conversational search.

BrowseSafe by Perplexity

Fine-tuned classifier detecting prompt-injection attacks in the Comet browser.

High confidence

Receipts · 2

Show quotes (2)
  • “fine-tuned a version of Qwen3-30B that can scan raw HTML and detect prompt injection attacks” — x.com, Dec 3, 2025
  • “BrowseSafe is designed to analyze complete webpages quickly without affecting browser performance.” — mpost.io, Dec 4, 2025

Source last verified Jul 26, 2026

Trail of Bits security audit by Trail of Bits

External security firm stress-tests Comet's agentic browsing against prompt injection.

High confidence

Receipts · 1

Show quotes (1)
  • “Before launching their Comet browser, Perplexity hired us to test the security of their AI-powered browsing features.” — blog.trailofbits.com, Feb 20, 2026

Source last verified Jul 26, 2026

Quora

Question-and-answer platform and operator of Poe, a multi-model AI chat app.

Modal Sandboxes by Modal

Securely executes LLM-generated code in Poe.

High confidence

Receipts · 1

Show quotes (1)
  • “Quora uses Modal Sandboxes to securely execute LLM-generated code in Poe, their AI chatbot platform.” — modal.com, Jun 30, 2025

Source last verified Jul 27, 2026

Shopify

E-commerce platform powering millions of merchant storefronts and an AI merchant assistant (Sidekick).

Refusal-behavior training pipeline by Shopify (in-house)

Trains Sidekick to refuse impossible or out-of-scope requests.

High confidence

Receipts · 1

Show quotes (1)
  • “The model had never learned to say no.” — shopify.engineering, Jun 15, 2026

Source last verified Jul 26, 2026

Stably

AI-powered QA testing agents for web applications.

Vercel Sandbox by Vercel

Enables secure agentic code execution.

High confidence

Receipts · 1

Show quotes (1)
  • “and Sandbox for agentic code execution.” — vercel.com, Feb 17, 2026

Source last verified Jul 27, 2026