Enforces fine-grained, context-aware access control on MCP tool calls, with a tamper-evident, replayable audit log that records denials and verifies every decision.
A fail-closed cryptographic gate for the MCP tool-call boundary that intercepts tools/call requests, evaluates a policy, and either forwards or denies the call with signed receipts, providing tamper-evident evidence for AI agent actions.
Enables an MCP server to gate tools/call requests by requiring a per-send channel-policy attestation for outbound actions, checking ROPE-lite origin tags on sensitive arguments, and returning SEP-2643-shaped structured denials with retry handles and remediation hints when a call is blocked. It logs client-asserted audit context (such as turnId and userIntent) for correlation across allowed and denied calls, but never uses that receipt for allow/deny decisions.
Enables MCP-speaking agents to execute tool calls through an authorization gate, producing signed decision records and reconciled effect rows for accountability.