---
name: aisecurityguard.io
description: aisecurityguard.io provides a single-endpoint preflight validation service that checks individual values — URLs, prices, integers, addresses, and hashes — for structural correctness and potential security issues before they are passed to downstream systems. It returns a verdict, confidence score, and flags describing any detected anomalies.
host: aisecurityguard.io
---

# aisecurityguard.io

aisecurityguard.io is a lightweight input-validation host aimed at agents that need a fast, structured security check on a single value before acting on it. It is not a full security platform; it handles one value per call and focuses on structural and surface-level security signals rather than deep semantic or behavioral analysis.

## When to use this host

Use this host when an agent needs a quick, structured sanity-and-security check on a single value before passing it to a downstream system — for example, verifying a wallet address looks valid before submitting a transaction, or checking a URL for suspicious patterns before fetching it. Do not use it for batch validation of multiple fields in one call, full AI prompt-injection analysis, SQL injection detection, or any deep behavioral or semantic security analysis. For those needs, look for a dedicated prompt-security or WAF-style host. Because this host validates only one value per call, agents validating many fields in a single payload will need to make multiple sequential calls.

## Capabilities

### Preflight Input Validation

Validates a single value of a declared type (URL, price, integer, address, or hash) for structural correctness and security red flags, returning a verdict, suspicious flag, and detailed flags array.

- **`validate-preflight-input`** — Performs a lightweight structural security validation on a single value (URL, price, integer, address, or hash) and returns a verdict with flags and confidence score.

## Skill reference

### `validate-preflight-input`

**AIGuard Preflight Validation** — Performs a lightweight structural security validation on a single value (URL, price, integer, address, or hash) and returns a verdict with flags and confidence score.

*Use when:* Use when an agent needs to validate a URL, price, integer, address, or hash for structural correctness and security issues before passing it to a downstream system or executing an action.

*Not for:* Do not use for full AI prompt injection analysis or multi-field batch validation; this endpoint validates a single value per call and does not evaluate SQL injection strings or chain/decimals fields beyond their declared type.

**Inputs:**

- `type` (string, required) — The type of value to validate. One of: url, price, integer, address, hash.
- `value` (string, required) — The value to validate, as a string. For URLs, the full URL string; for addresses, the address string, etc.
- `chain` (string) — Blockchain context for the validation (e.g. evm). Used when validating addresses or hashes on a specific chain.
- `input` (string) — Additional raw input string to include in the validation context.
- `decimals` (integer) — Number of decimal places, relevant for price or numeric validations.
- `int_type` (string) — Integer type for bounds validation (e.g. uint256).
- `hash_type` (string) — Hash algorithm type for format validation (e.g. sha256).

**Returns:** Returns a result object with verdict=suspicious, valid=true, flags=["javascript_uri:to"], confidence=0.85, and details containing domain and scheme for the validated URL.

**Example:** `{"type": "url", "value": "https://api.malicious-site.com/redirect?to=javascript%3Aalert(1)", "chain": "evm"}`

---