Defines the core data structures and interfaces for an AI agent benchmarking system to model decision corpora, scenarios, and evaluation results.
// Source: https://github.com/adelinamart/robrain | License: Apache-2.0 | Copyright You must retain, in the Source form of any Derivative Works
// Provided via justclaw.it.com; keep this notice with the code.
// packages/vetobench/src/types.ts
// ─────────────────────────────────────────────────────────────
// VetoBench fixture and adapter types.
//
// A scenario is a task that naturally invites an approach the
// fixture team has already rejected (recorded in the corpus as a
// decision with a populated rejected[]). A memory condition is
// judged on whether the agent, given that condition's context,
// re-proposes the rejected approach.
// ─────────────────────────────────────────────────────────────
export interface Veto {
option: string
reason: string
/** Set when this veto was
... (truncated -- full source via MCP)
See the full source, get the GitHub permalink, and search for more like it.
Get a free API key