Frequently asked questions
blocksign.red is a blockchain-anchored e-signature platform built for humans and AI agents. Answers below cover legality, agent integration, payment, and verification.
What is blocksign.red?
blocksign.red is a blockchain-anchored electronic signature platform. Every signed agreement is sealed to two public blockchains — the XRP Ledger (confirmed in 2-3 seconds) and Bitcoin via OpenTimestamps — producing a tamper-evident audit trail that anyone can verify forever, without trusting blocksign.
Are blocksign signatures legally binding?
Yes. blocksign.red signatures satisfy the US ESIGN Act and UETA, and support eIDAS SES/AES/QES levels in the EU and UK. Under UETA §14, contracts formed by electronic agents are enforceable, so agreements initiated by AI agents on blocksign are legally valid.
How does an AI agent send a contract for e-signature?
An AI agent calls POST /v1/agent/agreements with a template name (nda, services-agreement, employment-offer, contractor, or sales-agreement) and variables, Markdown, or a base64 PDF — one call creates and sends the envelope. Agents can also connect over the Model Context Protocol at https://mcp.blocksign.red/mcp, which exposes tools like create_agreement, verify_agreement, and purchase_credits.
How do AI agents authenticate with blocksign?
Agents use scoped API keys with the bsk_agent_ prefix, sent as an Authorization: Bearer header. Keys are issued at blocksign.red/settings/api-keys or programmatically via POST /v1/api-keys. Capabilities (like billing.auto_purchase) restrict what each key can do.
Can an AI agent pay for documents without a human?
Yes. When payment is needed the API returns HTTP 402 with a machine-readable actions array: pay per document ($2.99 for up to 10 pages), buy a credit pack (10/$12, 50/$50, 200/$160), or start a subscription (Pro $29/mo, Business $79/mo). Agent keys with the billing.auto_purchase capability and a card on file are charged automatically and never see a 402.
What is blocksign's MCP server?
blocksign.red runs a Model Context Protocol server at https://mcp.blocksign.red/mcp (Streamable HTTP). It exposes 11 tools, 4 resources, and 3 guided prompts so MCP clients — Claude, ChatGPT, IDEs, and autonomous agents — can create, track, negotiate, and verify agreements directly. Discovery metadata lives at /.well-known/mcp.json and /.well-known/mcp/server-card.json.
How do I verify a signed document?
Anyone can verify a blocksign agreement with no account and no authentication: GET /v1/verify/{envelope_id} (or the /verify page) recomputes the audit-chain hashes and checks the XRP Ledger and Bitcoin anchors. Verification is independent — it works even if blocksign disappears.
Can two AI agents negotiate a contract on blocksign?
Yes. Each signer can specify an agent_mcp_url or agent_callback_url. The counterparty's agent receives a scoped workflow token (72-hour TTL) and can review the proposal, submit counter-proposals, and accept terms via the /v1/agent/agreements/{id}/counter and /accept endpoints.
Where do AI agents find blocksign's API documentation?
Machine-readable docs: /llms.txt (quick start), /llms-full.txt (complete reference), /v1/openapi.json (OpenAPI 3.1), /v1/agent/instructions (structured JSON bootstrap), and /.well-known/mcp.json (MCP discovery).