1"use strict";(globalThis.webpackChunkhex_docs||=[]).push([[3319],{508234(e,t,n){n.r(t),n.d(t,{default:()=>c,metadata:()=>s});var s=n(897364),a=n(474848),r=n(28453),o=n(171006);const i={title:"\ud83e\uddea\xa0Introducing Evals",description:"Measure how the Hex Agent performs and test context changes safely before they ship, all from the CLI.",slug:"2026-08-04",date:new Date("2026-08-04T15:00:00.000Z")};function l(e){const t={a:"a",h3:"h3",li:"li",p:"p",strong:"strong",ul:"ul",...(0,r.R)(),...e.components};return(0,a.jsxs)(a.Fragment,{children:[(0,a.jsx)(t.h3,{id:"introducing-evals",children:"\ud83e\uddea\xa0Introducing Evals"}),"\n",(0,a.jsx)(t.p,{children:"Measure Hex Agent performance and test context changes safely before they ship, all from the CLI."}),"\n",(0,a.jsx)(t.p,{children:"If you've added a guide or tweaked a semantic model, Evals helps you measure whether it actually helped \u2014 before you publish. To do that, define a suite of test questions with expected outcomes and run them against the Hex Agent."}),"\n",(0,a.jsx)(o.A,{src:"https://www.youtube.com/embed/v15LyyPUNV8?si=KRlWkXrExReDuWUM"}),"\n",(0,a.jsx)(t.p,{children:"Evals turn agent quality and cost into something that you can measure and improve, so you're not just going on vibes."}),"\n",(0,a.jsx)(t.p,{children:"Some ways teams are using it:"}),"\n",(0,a.jsxs)(t.ul,{children:["\n",(0,a.jsxs)(t.li,{children:[(0,a.jsx)(t.strong,{children:"Test context changes before publishing"})," by forking context and running Evals to validate agent improvements."]}),"\n",(0,a.jsxs)(t.li,{children:[(0,a.jsx)(t.strong,{children:"Catch drift before your users do"})," by running the same Evals against production on a cadence."]}),"\n",(0,a.jsxs)(t.li,{children:[(0,a.jsx)(t.strong,{children:"Run configuration sweeps"})," to compare cost and pass rate side by side."]}),"\n"]}),"\n",(0,a.jsxs)(t.p,{children:[(0,a.jsx)(t.a,{href:"https://hex.tech/blog/evals/",children:"Learn more about Evals"})," or ",(0,a.jsx)(t.a,{href:"/docs/agent-management/evals",children:"dive into the docs"})," for the full rubric reference."]}),"\n",(0,a.jsxs)(t.p,{children:[(0,a.jsx)(t.strong,{children:"Getting started:"})," ",(0,a.jsx)(t.a,{href:"/docs/api-integrations/cli",children:"Install the Hex CLI"})," and follow the bootstrap prompt for Claude Code, Cursor, or your coding agent of choice \u2014 it scans your recent Threads, clusters them into your team's most common question topics, and drafts a starter suite from real usage instead of a blank file."]})]})}function c(e={}){const{wrapper:t}={...(0,r.R)(),...e.components};return t?(0,a.jsx)(t,{...e,children:(0,a.jsx)(l,{...e})}):l(e)}n.d(t,["assets",0,{authorsImageUrls:[]},"contentTitle",0,void 0,"frontMatter",0,i,"toc",0,[{value:"\ud83e\uddea\xa0Introducing Evals",id:"introducing-evals",level:3}]])},897364(e){e.exports=JSON.parse('{"permalink":"/changelog/2026-08-04","source":"@site/changelog/2026-08-04.md","title":"\ud83e\uddea\xa0Introducing Evals","description":"Measure how the Hex Agent performs and test context changes safely before they ship, all from the CLI.","date":"2026-08-04T15:00:00.000Z","tags":[],"hasTruncateMarker":false,"authors":[],"frontMatter":{"title":"\ud83e\uddea\xa0Introducing Evals","description":"Measure how the Hex Agent performs and test context changes safely before they ship, all from the CLI.","slug":"2026-08-04","date":"2026-08-04T15:00:00.000Z"},"unlisted":false,"prevItem":{"title":"Tune how the agent runs","permalink":"/changelog/2026-08-10"},"nextItem":{"title":"Command (line) and conquer","permalink":"/changelog/2026-07-30"}}')}}]);
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.