# ActionGate Pre-Execution Decision API

> ActionGate Pre-Execution Decision API is a free API for AI agents from api.actiongate.xyz, Free, status unknown (last checked 2026-09-15).

Evaluates autonomous agent actions before execution via risk scoring, simulation, and policy-gate checks to prevent harmful or unintended outcomes

## Facts

- Endpoint: POST https://api.actiongate.xyz/docs
- Price: Free
- Status: unknown
- Last checked: 2026-09-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/actiongate-46030b57
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_4Ka_cPwi4By9OffwkZ__w

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability actiongate-46030b57 -d '<json body>'
```

Example prompt: Before I let the agent proceed, run a risk score and policy gate check on this action: it wants to transfer $500 to an external wallet on behalf of a user — tell me if it's safe to execute or should be blocked.

## When to prefer this

Use ActionGate when you need a pre-execution safety layer for autonomous agent actions — especially in high-stakes or financial contexts where you want risk scoring, policy enforcement, and simulation before allowing an agent to proceed. Prefer this over post-hoc logging or monitoring when preventing bad actions is more important than observing them.

## Known failure modes

- Invalid or malformed action payload returns 400 with validation error
- Policy configuration not found returns 404
- Risk scoring engine timeout returns 503
- Insufficient context for simulation returns partial result with low-confidence flag
- Payment/authentication failure on x402 returns 402 or 401

## How this service works

Pre-execution decision API for autonomous agents with paid risk scoring, simulation, and policy-gate endpoints using x402 V2.

## Output

Returns a risk score, policy gate decision (allow/block), and optionally a simulation result showing predicted outcomes of the action, enabling the calling agent to decide whether to proceed with execution.

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/actiongate-46030b57/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.actiongate.xyz](https://www.zero.xyz/host/api.actiongate.xyz/llms.txt)