---
name: api.zuluworksai.com
description: api.zuluworksai.com exposes two independent API skills: one that scores the trustworthiness of the calling agent's own connection using TLS, HTTP, geo, and ASN signals, and one that runs a text prompt through Cloudflare Workers AI's hosted Llama 3.1 8B instruct model. Both are billed per-use and return structured JSON responses.
host: api.zuluworksai.com
---

# api.zuluworksai.com

ZuluWorks AI is a small merchant host offering two distinct, pay-per-use capabilities: a connection trust-scoring endpoint (KYA) and a hosted LLM inference endpoint. It targets agents that need lightweight, on-demand access to either connection hygiene checks or Llama 3.1 8B text generation without managing their own infrastructure. The two skills are independent and serve different use cases.

## When to use this host

Use this host when an agent needs a quick, per-use connection trust check against its own outbound properties (not arbitrary third-party URLs), or when it needs single-turn Llama 3.1 8B text completions via a simple GET request without managing model infrastructure. Do not use fetch-kya-trust-score to score external domains you supply as input — it only reflects the calling connection's own properties. Do not use fetch-workers-ai-llama-completion for streaming output, multi-turn conversations with history, or models other than llama-3.1-8b-instruct; for those needs, route to a dedicated LLM provider (e.g., OpenAI, Anthropic, or a Cloudflare Workers AI integration with broader model support).

## Capabilities

### Connection Trust Scoring

Evaluates the calling agent's own connection at request time by inspecting TLS version, HTTP headers, IP resolution, geographic origin, and ASN, returning a structured trust score and signal breakdown.

- **`fetch-kya-trust-score`** — Computes a live agent handshake trust score from TLS, HTTP protocol, user-agent, geo, ASN, and routing signals for the calling client.

### LLM Text Completion

Accepts a text prompt and returns a generated response from the Cloudflare-hosted Llama 3.1 8B instruct model, billed per call at $0.002 USDC on Base.

- **`fetch-workers-ai-llama-completion`** — Runs a text prompt through Cloudflare Workers AI using the @cf/meta/llama-3.1-8b-instruct model and returns the generated response.

## Workflows

### Trust-Gated LLM Inference

*Use when an agent needs to verify its own connection meets a minimum trust threshold before submitting a prompt to the LLM endpoint, ensuring sensitive inference requests originate from adequately secured connections.*

1. **`fetch-kya-trust-score`** — Score the calling agent's current connection; inspect trust_score and signals to determine whether the connection meets the required threshold.
2. **`fetch-workers-ai-llama-completion`** — If the trust score passes the threshold, proceed to submit the prompt and retrieve the Llama 3.1 8B generated response.

## Skill reference

### `fetch-kya-trust-score`

**KYA Trust Score** — Computes a live agent handshake trust score from TLS, HTTP protocol, user-agent, geo, ASN, and routing signals for the calling client.

*Use when:* Use when an agent needs to verify the trustworthiness of an inbound or outbound connection by scoring its TLS version, HTTP headers, IP resolution, geographic origin, and ASN before proceeding with a sensitive operation.

*Not for:* Do not use for scoring arbitrary third-party URLs or domains you supply as input; this endpoint scores the calling agent's own connection properties at request time.

**Returns:** Returns a trust_score (e.g. 0.92), handshake_verification status, per-signal breakdown (TLS, geo, ASN, cf_edge, etc.), request_id, and subscription quota details.

**Example:** `GET https://api.zuluworksai.com/api/kya`

---

### `fetch-workers-ai-llama-completion`

**Workers AI Llama Completion** — Runs a text prompt through Cloudflare Workers AI using the @cf/meta/llama-3.1-8b-instruct model and returns the generated response.

*Use when:* Use when an agent needs to generate a text completion or answer from a Llama 3.1 8B instruct model via a simple GET request, billed per-use at $0.002 USDC on Base.

*Not for:* Do not use for streaming completions, multi-turn chat sessions requiring conversation history, or tasks requiring a model other than llama-3.1-8b-instruct.

**Inputs:**

- `prompt` (string, required) — The text prompt to send to the Llama model for inference.

**Returns:** Returns a JSON object with the model's text response under fulfillment.response, the model identifier, echo of the prompt, and subscription quota details including remaining requests and expiry.

**Example:** `GET https://api.zuluworksai.com/api/workers-ai?prompt=Explain+post-quantum+cryptography+in+one+paragraph.`

---