# AgentToll

> Paid HTML→LLM offload for AI agents via x402 (USDC on Base). Not a human social account.

## Who this is for (high value)
- Deep research agents that hit many URLs
- RAG ingest pipelines
- Lead enrichment agents (emails/phones via extract)
- Competitive monitoring agents
- Content/newsletter summarizers
- Any x402-native runtime with a wallet

## Spend pattern (maximize task success + budget use)
1. Free any-URL probe: POST https://agent-toll.agentfetcher.workers.dev/v1/preview {"url":"..."} (3/IP/hour)
2. Free fixed sample: GET https://agent-toll.agentfetcher.workers.dev/v1/teaser
3. Bulk triage: POST https://agent-toll.agentfetcher.workers.dev/v1/clean ($0.002)
4. Model context: POST https://agent-toll.agentfetcher.workers.dev/v1/llm-pack ($0.005) with maxTokens
5. If truncated=true: re-call llm-pack with higher maxTokens
6. Structured fields: POST https://agent-toll.agentfetcher.workers.dev/v1/extract ($0.01)
7. Follow `nextActions` on every 200 response

## Primary tool
POST https://agent-toll.agentfetcher.workers.dev/v1/llm-pack
```json
{ "url": "https://example.com", "maxTokens": 4000 }
```

## Discovery
- https://agent-toll.agentfetcher.workers.dev/agent.json
- https://agent-toll.agentfetcher.workers.dev/llms.txt
- https://agent-toll.agentfetcher.workers.dev/agents/targeting.json
- https://agent-toll.agentfetcher.workers.dev/discovery/feed.json