TrustNotch provides tamper-evident audit infrastructure designed specifically for AI agents, addressing the need for trustworthy and independently verifiable logs of agent actions. The platform ensures that every action submitted by an AI agent is logged with cryptographic integrity and anchored for long-term verifiability, even if the service itself is no longer available.
When an agent submits an action to TrustNotch over HTTP, the system returns an Ed25519-signed receipt, which serves as a cryptographic commitment to the precise content of the logged action. These log entries are then batched into a Merkle tree following RFC 6962, with each batch root anchored to the Bitcoin blockchain using OpenTimestamps. This process provides each entry with a tamper-evident timestamp and links it to a widely audited, decentralized ledger. As a result, the integrity and existence of each log entry can be proven independently of TrustNotch.
Verification of log entries does not require ongoing reliance on the TrustNotch service. Anyone can use the open-source TrustNotch verifier, available as a standalone package, to check the validity of receipts, Merkle tree inclusion, and Bitcoin anchoring entirely offline. This design allows proofs of logged actions to remain verifiable even if TrustNotch ceases operation.
TrustNotch offers multiple integration points for users and developers. Its MCP Server can connect AI agents such as Claude Desktop and other MCP clients, with the server package installable via pip or runnable with uvx. com for submitting, fetching, and proving log entries. Both the verifier and the MCP server are available as Python packages on PyPI. TrustNotch is built and hosted in the EU, specifically in Helsinki, Finland.
This tool is positioned as audit infrastructure for AI, providing cryptographically signed, Bitcoin-anchored, and independently verifiable audit logs for agents operating in sensitive or regulated environments.
TrustNotch sits in PulseGate's Infrastructure & Backend category. It focuses on ensuring the integrity and verifiability of audit logs for AI agent actions. It is built as a B2B product for ai infrastructure developers. It runs on the command line and API.
TrustNotch builds and maintains TrustNotch, and the product first shipped in 2026. PulseGate's similarity index finds few close equivalents — TrustNotch occupies a relatively distinct niche. Key capabilities include tamper-evident logs, bitcoin anchoring, and offline verification. It exposes integrations via an MCP server and a public API.
Latest indexed changes and source events
trustnotch.com discovered by the PulseGate indexer
Other apps tracked under the same category.
