# Sloopt

> Sloopt (formerly SkillGrade) independently tests how AI agents and agent skills behave under attack and grades them A to F, with the evidence for every finding, mapped to OWASP, MITRE ATLAS, NIST AI RMF and the EU AI Act. Two products: Rate (grade a skill before you install it: free check, $79 certification at 32 runs, $29/month monitoring) and Guard (audit an agent you already run: read-only repo access plus live probing of the running agent, one-off report from $500, subscription from $199/month). Grades come only from real engine runs and are never edited by hand. A grade is a measurement at a point in time, not a safety guarantee. Website: https://sloopt.ai · Support: support@sloopt.ai

## Product
- [Rate — grade a skill before you install it](https://sloopt.ai/rate): paste a SKILL.md or a link; seven behavioral controls; injection probes run 8 times free, 32 times for certification; worst finding sets the letter.
- [Guard — audit an agent you already run](https://sloopt.ai/guard): read-only GitHub App on the repo plus the URL of the running agent; static scanners and live injection probes (NVIDIA Garak); A–F findings report with file, line, fix and standards clause.
- [Pricing](https://sloopt.ai/pricing): Rate free / $79 / $29 per month. Guard from $500 once or $199 per month. No credits, no metered bills. Payments via Creem.
- [For teams and enterprises](https://sloopt.ai/enterprise)

## How grading works
- Agents and tools should request a grade via POST /check (the grade API), not by opening checkout. An MCP server for the same call is tracked separately.
- [Methodology](https://sloopt.ai/methodology): the seven controls, run counts, statistical bounds, how the letter is set, published accuracy (agreement with expert labels, false-positive and false-negative rates).
- [Compliance and standards mapping](https://sloopt.ai/compliance): OWASP Top 10 for LLM Applications, OWASP Top 10 for Agentic Applications (ASI01–ASI10), MITRE ATLAS, NIST AI RMF, EU AI Act, ISO 42001, SOC 2.
- [Rubric changelog](https://sloopt.ai/changelog)
- [Disputes and appeals](https://sloopt.ai/disputes)

## Graded skills (public, permanent pages)
- [All graded skills](https://sloopt.ai/skills)
- [Is ui-ux-pro-max safe? Grade C](https://sloopt.ai/skill/ui-ux-pro-max-9a03bd2e): tested 2026-10-07.
- [Is ui-ux-pro-max safe? Grade C](https://sloopt.ai/skill/ui-ux-pro-max-748b6303): tested 2026-10-07.
- [Is product?id=4b54489a-c5e6-4432-93ff-96caa4dec7ae safe? Grade A](https://sloopt.ai/skill/product-id-4b54489a-c5e6-4432-93ff-96caa4dec7ae-b4ef527b): tested 2026-10-07.
- [Is self-improving agent safe? Grade A](https://sloopt.ai/skill/self-improving-agent-eff88995): tested 2026-08-06.
- [Is Browser Automation safe? Grade A](https://sloopt.ai/skill/browser-automation-9705b3b7): tested 2026-08-06.
- [Is Self Improving Agent safe? Grade A](https://sloopt.ai/skill/self-improving-agent-c4b1d7cc): tested 2026-08-06.
- [Is News Summary safe? Grade A](https://sloopt.ai/skill/news-summary-1b4c45b5): tested 2026-08-06.
- [Is ATXP safe? Grade A](https://sloopt.ai/skill/atxp-9876fb6c): tested 2026-08-06.
- [Is imap-smtp-email safe? Grade A](https://sloopt.ai/skill/imap-smtp-email-7bc652cd): tested 2026-08-06.
- [Is Himalaya safe? Grade A](https://sloopt.ai/skill/himalaya-2265054d): tested 2026-08-06.
- [Is Slack safe? Grade A](https://sloopt.ai/skill/slack-e15de2ab): tested 2026-08-06.
- [Is Powerpoint / PPTX safe? Grade A](https://sloopt.ai/skill/powerpoint-pptx-d286c8cd): tested 2026-08-06.
- [Is YouTube Watcher safe? Grade A](https://sloopt.ai/skill/youtube-watcher-89b97c95): tested 2026-08-06.
- [Is Video Frames safe? Grade A](https://sloopt.ai/skill/video-frames-7e242454): tested 2026-08-06.
- [Is Find Skills Skill safe? Grade A](https://sloopt.ai/skill/find-skills-skill-6a456ab4): tested 2026-08-06.
- [Is Stock Analysis safe? Grade A](https://sloopt.ai/skill/stock-analysis-6ae566bd): tested 2026-08-06.
- [Is Brave Search safe? Grade A](https://sloopt.ai/skill/brave-search-57ec3e0c): tested 2026-08-06.
- [Is Desktop Control safe? Grade A](https://sloopt.ai/skill/desktop-control-d900c065): tested 2026-08-06.
- [Is Elite Longterm Memory safe? Grade A](https://sloopt.ai/skill/elite-longterm-memory-56bb3416): tested 2026-08-06.
- [Is Mcporter safe? Grade A](https://sloopt.ai/skill/mcporter-47525316): tested 2026-08-06.
- [Is Free Ride - Unlimited free AI safe? Grade A](https://sloopt.ai/skill/free-ride-unlimited-free-ai-9e016001): tested 2026-08-06.
- [Is Feishu Evolver Wrapper safe? Grade A](https://sloopt.ai/skill/feishu-evolver-wrapper-f5d84210): tested 2026-08-06.
- [Is Agent Phone Call safe? Grade A](https://sloopt.ai/skill/agent-phone-call-8db52e6c): tested 2026-08-06.
- [Is Browser Use safe? Grade A](https://sloopt.ai/skill/browser-use-a8cdeb87): tested 2026-08-06.
- [Is Excel / XLSX safe? Grade A](https://sloopt.ai/skill/excel-xlsx-52421d9f): tested 2026-08-06.
- [Is Automation Workflows safe? Grade A](https://sloopt.ai/skill/automation-workflows-7e2a1b38): tested 2026-08-06.
- [Is Prismfy Web Search | Free Google safe? Grade A](https://sloopt.ai/skill/prismfy-web-search-free-google-43f0df0c): tested 2026-08-06.
- [Is Evolver safe? Grade A](https://sloopt.ai/skill/evolver-ff0c2e6f): tested 2026-08-06.
- [Is API Gateway safe? Grade A](https://sloopt.ai/skill/api-gateway-f2f3013b): tested 2026-08-06.
- [Is Sonoscli safe? Grade A](https://sloopt.ai/skill/sonoscli-0b9cf84b): tested 2026-08-07.
- [Is Word / DOCX safe? Grade A](https://sloopt.ai/skill/word-docx-c766da48): tested 2026-08-06.
- [Is Baidu web search safe? Grade A](https://sloopt.ai/skill/baidu-web-search-4fa6b504): tested 2026-08-06.
- [Is Auto-Updater Skill safe? Grade A](https://sloopt.ai/skill/auto-updater-skill-4142916e): tested 2026-08-06.
- [Is Skill Creator safe? Grade A](https://sloopt.ai/skill/skill-creator-b2e3d83f): tested 2026-08-06.
- [Is Notion safe? Grade A](https://sloopt.ai/skill/notion-1002bcb8): tested 2026-08-06.
- [Is PollyReach safe? Grade A](https://sloopt.ai/skill/pollyreach-bf71d4da): tested 2026-08-06.
- [Is Tavily 搜索 safe? Grade A](https://sloopt.ai/skill/tavily-490da2ad): tested 2026-08-06.
- [Is Nano Banana Pro safe? Grade A](https://sloopt.ai/skill/nano-banana-pro-4ea88f4b): tested 2026-08-08.
- [Is Obsidian safe? Grade A](https://sloopt.ai/skill/obsidian-dc45b522): tested 2026-08-06.
- [Is Nano Pdf safe? Grade A](https://sloopt.ai/skill/nano-pdf-4a19c391): tested 2026-08-06.
- [Is Humanizer safe? Grade A](https://sloopt.ai/skill/humanizer-d7ce483d): tested 2026-08-06.
- [Is AdMapix safe? Grade A](https://sloopt.ai/skill/admapix-9ce58533): tested 2026-08-06.
- [Is Agent Browser safe? Grade A](https://sloopt.ai/skill/agent-browser-9643b9c6): tested 2026-08-06.
- [Is Multi Search Engine safe? Grade A](https://sloopt.ai/skill/multi-search-engine-a089b792): tested 2026-08-06.
- [Is Weather safe? Grade A](https://sloopt.ai/skill/weather-1ca0c8d7): tested 2026-08-06.
- [Is Proactive Agent safe? Grade A](https://sloopt.ai/skill/proactive-agent-571521d9): tested 2026-08-06.
- [Is SkillScan safe? Grade A](https://sloopt.ai/skill/skillscan-7cc318d8): tested 2026-08-06.
- [Is Gog safe? Grade A](https://sloopt.ai/skill/gog-00c1af6c): tested 2026-08-06.
- [Is Github safe? Grade A](https://sloopt.ai/skill/github-51b28818): tested 2026-08-06.
- [Is ontology safe? Grade A](https://sloopt.ai/skill/ontology-669830c5): tested 2026-08-06.
- [Is Self-Improving + Proactive Agent safe? Grade A](https://sloopt.ai/skill/self-improving-proactive-agent-11109b39): tested 2026-08-06.
- [Is Skill Vetter safe? Grade A](https://sloopt.ai/skill/skill-vetter-e8eb7583): tested 2026-08-06.
- [Is self-improving agent safe? Grade A](https://sloopt.ai/skill/self-improving-agent-d27e7def): tested 2026-08-08.

## Answers
- [Answers index](https://sloopt.ai/answers)
- [Tools to test AI agent safety in 2026: garak, promptfoo, mcp-scan, AIRQ, Sloopt and others — what each actually measures](https://sloopt.ai/answers/ai-agent-security-testing-tools-compared): AI agent security testing tools compared: garak, promptfoo, MCP scanners, AIRQ and Sloopt, by what each measures and what it leaves out.
- [What evidence does the EU AI Act expect for an AI agent, and what a buyer's questionnaire will ask](https://sloopt.ai/answers/eu-ai-act-evidence-for-ai-agents): EU AI Act requirements turned into the evidence an AI agent vendor needs on file, plus how to answer the 15 AI questions in a security questionnaire.
- [How to test an AI agent for prompt injection (and why one run proves nothing)](https://sloopt.ai/answers/how-to-test-an-ai-agent-for-prompt-injection): Step-by-step prompt-injection testing for AI agents and skills: run counts, statistical bounds and the real pass rates from Sloopt's engine runs.
- [Is this AI agent skill safe to install? A 10-minute check, and what 47 graded skills showed](https://sloopt.ai/answers/is-this-ai-agent-skill-safe-to-install): How to check whether an AI agent skill (SKILL.md) is safe before you install it: manual checks, behavioural grading, and what 47 graded skills showed.
- [OWASP Top 10 for Agentic Applications (ASI01–ASI10): a checklist you can actually test](https://sloopt.ai/answers/owasp-top-10-agentic-applications-checklist): Every OWASP agentic risk, ASI01 to ASI10, with a concrete test, the evidence to keep, and what an automated grader can and cannot assess today.

## Company
- [About Sloopt, formerly SkillGrade](https://sloopt.ai/about)
- [Security and your data](https://sloopt.ai/security)
- [Terms](https://sloopt.ai/terms) · [Privacy](https://sloopt.ai/privacy) · [Refunds](https://sloopt.ai/refunds) · [Acceptable use](https://sloopt.ai/acceptable-use)

## Optional
- [State of AI Skill Security report — sign up](https://sloopt.ai/report)