#hashline
v0.1 · early access

Tamper-evident audit logs
for AI agents.

Every prompt, tool call, and decision — hash-chained and verifiable. Know exactly what your agent did, and prove it.

Request early accessView SDK on GitHub →
agent.tsTypeScript
import { Client } from "@hashline/sdk";

const hl = new Client({{ apiKey: process.env.HASHLINE_API_KEY }});

await hl.event({{
  run_id: "run_01HXYZ...",
  type: "tool_call",
  actor: {{ type: "agent", id: "search_agent" }},
  payload: {{
    name: "web_search",
    arguments: {{ q: "hashline" }},
    call_id: "c1",
  }},
}});
Why Hashline

Hash-chained integrity

Every event is cryptographically chained to the one before it. Tampering with any record breaks the chain — detectable instantly via a single API call.

Built for agents, not apps

Traditional logging assumes one app serving many users. Agents are one-to-one: unique runs, dynamic tool calls, unpredictable paths. Hashline is structured around runs, not requests.

Compliance-ready exports

Export any run as a signed, verifiable evidence pack. Hand it to an auditor, a regulator, or your own security team. The chain proves nothing was altered after the fact.

Pricing

Transparent tiers. No “contact us” for what a simple product costs.

Hobby
Free
  • Events100K / mo
  • Retention7 days
  • ExportJSONL
Get started
Most popular
Pro
$29/mo
  • Events10M / mo
  • Retention90 days
  • ExportJSONL
Request access
Team
$199/mo
  • Events100M / mo
  • Retention1 year
  • ExportCompliance pack
Request access

Need more? → hello@hashline.dev