Skip to content
Agent Provenance LedgerAutonomous AI Governance

Immutable cryptographic audit trails for AI-generated outbound emails

When autonomous AI agents send emails on your behalf, prove what was sent, why it was generated, and which prompt triggered it with tamper-proof SHA-256 provenance headers and Ed25519 signatures. Includes 6,000 free monthly emails.

Live Agent Provenance Inspector

Simulate cryptographic SHA-256 audit trails and Ed25519 signature binding across autonomous agent triggers.

SHA-256 Cryptographically Verified
1. Origin & Agent Execution Lineage
Canonical Agent Name:claude-code-cli/v1.2
Agent Role / Purpose:Autonomous Devops Agent
Trigger Event:git_tag_push
Intended Recipient:engineering-alerts@acme.com
Subject & Payload Summary:

Production Deployment Complete: v2.4.1 (Clean 0 Errors)

Automated release summary generated following 100% test pass on main branch.

2. Cryptographic Attestation & Headers
X-SadaSend-Prompt-Hash (SHA-256):
e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
X-SadaSend-Tool-Call-Id:
call_deploy_notify_98214
X-SadaSend-Envelope-Digest:
7f83b1657ff1fc53b92dc18148a1d65dfc2d4b1fa3d677284addd200126d9069
X-SadaSend-Provenance-Sig (Ed25519):
sig_ed25519_9d81a4ef882bc1947e301d0a5197822fb142a5bc7719f9
Core Architecture

Four layers of cryptographic proof for autonomous dispatch

Input Attribution

SHA-256 Prompt Fingerprinting

Every agent send binds a cryptographic SHA-256 digest of the triggering prompt and system context, preventing unverified prompt injection claims.

Execution Lineage

Tool Call ID Binding

Directly links the email envelope to the autonomous agent’s unique tool call identifier, establishing an unbroken execution chain.

Ed25519 Integrity

Tamper-Proof Envelope Signatures

Outbound MIME bodies and recipient headers are cryptographically signed at ingestion, proving zero payload tampering in the delivery relay.

SOC2 & ISO 27001

Instant Enterprise Compliance

Provides security and compliance officers with complete, exportable provenance records resolving the AI agent accountability gap.

SDK Integration

Drop-in provenance tracking across modern agent stacks

TYPESCRIPT
import { SadaSend } from 'sadasend';

const sadasend = new SadaSend({ apiKey: process.env.SADASEND_API_KEY! });

// Dispatch with complete agent provenance attribution
const result = await sadasend.emails.send({
  from: 'alerts@yourdomain.com',
  to: 'compliance@client.com',
  subject: 'Automated Security Audit Report',
  html: '<p>Attached: Monthly vulnerability remediations verified by agent.</p>',
  provenance: {
    agentName: 'claude-code-cli/v1.2',
    prompt: userPromptText, // Automatically hashed to SHA-256 in flight
    toolCallId: 'call_audit_98124',
    metadata: {
      framework: 'claude-code',
      executionCycle: 'nightly-ci-eval',
    },
  },
});

console.log('Dispatched with cryptographic provenance:', result.provenanceDigest);
Governance & Audit

Frequently asked questions about agent provenance

Give your autonomous agents accountable, tamper-proof email delivery

Start free with 6,000 emails every month. Full cryptographic provenance headers, SpendGuard \$0 floor, and Dual-DKIM signing included.