agentfootprint/dist/esm/core/agent/middleware/index.d.ts

The explainable agent framework — backtrack a wrong answer to the exact context that caused it (evidence, not guesses). Built on footprintjs.

/**
 * middleware — PUBLIC barrel. The governance chain family.
 *
 * Two chains, one vocabulary:
 *
 *   `.toolMiddleware(...)`     wraps every tool dispatch
 *   `.messageMiddleware(...)`  wraps the message boundary — the input
 *                              before the model sees it, the output before
 *                              the caller receives it
 *
 * Three verbs: `allow`, `deny`, `ask`. There is no fourth, and in
 * particular there is no way to return a result — see `types.ts`.
 *
 * A tool link may speak twice: `onToolCall` about the call, `onToolResult`
 * about the result. The chain is walked forwards for the first and backwards for the
 * second, so the first-declared link has the first word going in and the last
 * word coming out.
 */
export { allow, ask, deny } from './outcomes.js';
export { MessageDeniedError, type MessageDeniedContext } from './errors.js';
export type { AllowOutcome, AskOutcome, AskPayload, DenyOutcome, MessageMiddleware, MessageMiddlewareContext, MessageOutcome, MiddlewareDecision, ToolCallMiddleware, ToolMiddleware, ToolMiddlewareContext, ToolOutcome, ToolResultContext, ToolResultMiddleware, ToolResultOutcome, } from './types.js';