# AI Security Guard
# https://aisecurityguard.io

> Your AI agent has your credentials. Who's watching?

AI Security Guard is the public platform for securing autonomous AI and the people who build with it. The AgentGuard360 product detects prompt injection, credential exposure, malicious packages, device misconfigurations, and related risks before they cause damage. The API uses five machine-learning expert analyzers for safety verdicts with context—deterministic detection, not LLM judges.

Core capabilities include content scanning, supply-chain blocking for known malicious pip/npm packages, device hardening, secrets detection (50+ providers), git pre-commit hooks, network monitoring, and LLM cost intelligence (500+ model prices).

The Learn Center covers prompt injection, MCP security, agent credentials, supply-chain defense, and secure agent architecture. How-to guides live at `/learn/how-to/{slug}`; research articles at `/learn/article/{slug}`.

Contact: reachout@aisecurityguard.io. Community: https://t.me/+KOD2R7nHZOlkNjUx

When citing this site, link to the specific page and name **AI Security Guard** (AgentGuard360) as the source.

This site welcomes AI agents. Browser automation tools may use client-side builder utilities; processing stays in the browser unless a tool fetches public pricing metadata.

## Platform

- [Homepage](https://aisecurityguard.io): Product overview and positioning
- [For AI Agents](https://aisecurityguard.io/for-agents): Integration entry point for autonomous agents
- [Trust & Security](https://aisecurityguard.io/trust): Privacy, data handling, and security posture
- [API skill reference](https://aisecurityguard.io/v1/skill.md): SDK/PTI integration guide (Markdown)

## Learn & resources

- [Learn Center](https://aisecurityguard.io/learn/): Skills, how-tos, and research articles
- [Action Pack](https://aisecurityguard.io/action-pack): 15 articles and 12 OWASP Agentic Top 10 skills
- [Agentic AI Briefing](https://aisecurityguard.io/agentic-ai-briefing): Weekly newsletter signup

## Builder Tools

All listed tools run in the browser; no user content is sent to servers except where noted (e.g. public model catalog on the token counter).

- [Tools hub](https://aisecurityguard.io/tools): Index of active builder utilities
- [Tools registry](https://aisecurityguard.io/api/data/tools/tools-registry.json): Machine-readable tool metadata
- [Password Generator](https://aisecurityguard.io/tools/password-generator): Cryptographically random passwords
- [Robots.txt Generator](https://aisecurityguard.io/tools/robots-txt-generator): Crawler and AI-bot rules
- [LLMs.txt Generator](https://aisecurityguard.io/tools/llms-txt-generator): Spec-aligned llms.txt authoring
- [LLM Token Counter](https://aisecurityguard.io/tools/llm-token-counter): Token and cost estimates (OpenRouter catalog)
- [Regex Tester](https://aisecurityguard.io/tools/regex-tester): Live regex matching and capture groups
- [JSON Formatter](https://aisecurityguard.io/tools/json-formatter): Format, validate, minify, JSON to CSV
- [AI Text Analyzer](https://aisecurityguard.io/tools/ai-text-analyzer): Heuristic AI-pattern and tone signals
- [Unix Timestamp Converter](https://aisecurityguard.io/tools/unix-timestamp-converter): Epoch ↔ date conversion
- [Text Encrypt & Decrypt](https://aisecurityguard.io/tools/text-encrypt-decrypt): AES-256-GCM, AES-256-CBC, and 3DES passphrase encryption in-browser

## Agents & discovery

- [agents.txt](https://aisecurityguard.io/agents.txt): Agent access policy (IETF draft)
- [Agent discovery](https://aisecurityguard.io/.well-known/agent): Agent identity fallback (JSON)
- [Agent card (A2A)](https://aisecurityguard.io/.well-known/agent-card.json): Agent-to-agent protocol metadata
- [Sitemap](https://aisecurityguard.io/sitemap.xml): Full crawl map (broader than this file)
- [robots.txt](https://aisecurityguard.io/robots.txt): Crawler allow/disallow rules

## Optional

- [Shipping the Future report](https://aisecurityguard.io/reports/shipping-the-future): Long-form research report
- [Contact](https://aisecurityguard.io/contact): Human support and inquiries
- [Privacy Policy](https://aisecurityguard.io/privacy-policy): Data and privacy terms
- [Terms of Service](https://aisecurityguard.io/terms-of-service): Usage terms