@cursor/july 0.1.105 → 0.1.107

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (217) hide show
  1. package/dist/bin/agent-serve.js +1 -8
  2. package/dist/channels/bitbucket/types.d.ts +2 -4
  3. package/dist/channels/bitbucket/types.d.ts.map +1 -1
  4. package/dist/channels/deployments/types.d.ts +5 -8
  5. package/dist/channels/deployments/types.d.ts.map +1 -1
  6. package/dist/channels/github/types.d.ts +2 -4
  7. package/dist/channels/github/types.d.ts.map +1 -1
  8. package/dist/channels/gitlab/types.d.ts +2 -5
  9. package/dist/channels/gitlab/types.d.ts.map +1 -1
  10. package/dist/channels/origin/types.d.ts +2 -3
  11. package/dist/channels/origin/types.d.ts.map +1 -1
  12. package/dist/channels/slack/defaults.d.ts.map +1 -1
  13. package/dist/channels/slack/defaults.js +3 -0
  14. package/dist/channels/slack/dispatch.d.ts.map +1 -1
  15. package/dist/channels/slack/dispatch.js +1 -1
  16. package/dist/channels/slack/live-delivery.d.ts.map +1 -1
  17. package/dist/channels/slack/live-delivery.js +48 -11
  18. package/dist/channels/slack/types.d.ts +4 -3
  19. package/dist/channels/slack/types.d.ts.map +1 -1
  20. package/dist/docs/404.html +3 -3
  21. package/dist/docs/ab.html +6 -6
  22. package/dist/docs/assets/{app.BVwaffdK.js → app.CtuckIiL.js} +1 -1
  23. package/dist/docs/assets/chunks/@localSearchIndexroot.DXXZxiMv.js +1 -0
  24. package/dist/docs/assets/chunks/{VPLocalSearchBox.WuWTZ1x8.js → VPLocalSearchBox.Bkkhnu8K.js} +2 -2
  25. package/dist/docs/assets/chunks/{theme.BYdI_6sF.js → theme.DhpIe0Pa.js} +2 -2
  26. package/dist/docs/assets/{guides_opentelemetry.md.COh9jFVg.js → guides_opentelemetry.md.SCRMskno.js} +3 -3
  27. package/dist/docs/assets/{guides_opentelemetry.md.COh9jFVg.lean.js → guides_opentelemetry.md.SCRMskno.lean.js} +1 -1
  28. package/dist/docs/assets/{reference_channels.md.DBGcaB7m.js → reference_channels.md.D-qTqwcq.js} +2 -2
  29. package/dist/docs/assets/{reference_channels.md.DBGcaB7m.lean.js → reference_channels.md.D-qTqwcq.lean.js} +1 -1
  30. package/dist/docs/assets/{reference_cli.md.C6Lo8SEd.js → reference_cli.md.Dm67hd2D.js} +1 -1
  31. package/dist/docs/assets/{skills_otel.md.CqhobH3B.js → skills_otel.md.B5Ddv1xH.js} +1 -1
  32. package/dist/docs/assets/{style.BRuM8477.css → style.DCD_zyg3.css} +1 -1
  33. package/dist/docs/building-with-agents.html +6 -6
  34. package/dist/docs/concepts.html +5 -5
  35. package/dist/docs/deployment.html +6 -6
  36. package/dist/docs/evals.html +6 -6
  37. package/dist/docs/guides/agent-to-agent.html +6 -6
  38. package/dist/docs/guides/cloud-agents.html +6 -6
  39. package/dist/docs/guides/convert-automation.html +6 -6
  40. package/dist/docs/guides/github.html +6 -6
  41. package/dist/docs/guides/grokbot-agents.html +6 -6
  42. package/dist/docs/guides/human-in-the-loop.html +6 -6
  43. package/dist/docs/guides/mcp-oauth.html +6 -6
  44. package/dist/docs/guides/opentelemetry.html +8 -8
  45. package/dist/docs/guides/opentelemetry.md +3 -4
  46. package/dist/docs/guides/slack.html +6 -6
  47. package/dist/docs/guides/webhooks.html +6 -6
  48. package/dist/docs/hashmap.json +1 -1
  49. package/dist/docs/hillclimbing.html +6 -6
  50. package/dist/docs/index.html +6 -6
  51. package/dist/docs/llms-full.txt +22 -28
  52. package/dist/docs/quickstart.html +6 -6
  53. package/dist/docs/reference/agent-config.html +6 -6
  54. package/dist/docs/reference/artifacts.html +6 -6
  55. package/dist/docs/reference/channels.html +7 -7
  56. package/dist/docs/reference/channels.md +11 -19
  57. package/dist/docs/reference/cli.html +8 -8
  58. package/dist/docs/reference/cli.md +5 -2
  59. package/dist/docs/reference/connections.html +6 -6
  60. package/dist/docs/reference/extensions.html +6 -6
  61. package/dist/docs/reference/hooks.html +6 -6
  62. package/dist/docs/reference/http-api.html +6 -6
  63. package/dist/docs/reference/instructions.html +6 -6
  64. package/dist/docs/reference/playground.html +5 -5
  65. package/dist/docs/reference/project-layout.html +6 -6
  66. package/dist/docs/reference/prompt.html +6 -6
  67. package/dist/docs/reference/result.html +6 -6
  68. package/dist/docs/reference/schedules.html +6 -6
  69. package/dist/docs/reference/sessions.html +5 -5
  70. package/dist/docs/reference/skills.html +6 -6
  71. package/dist/docs/reference/subagents.html +6 -6
  72. package/dist/docs/reference/tools.html +6 -6
  73. package/dist/docs/scaffolding-agents.html +5 -5
  74. package/dist/docs/skills/ab.html +6 -6
  75. package/dist/docs/skills/create-agent.html +6 -6
  76. package/dist/docs/skills/debug.html +5 -5
  77. package/dist/docs/skills/evals.html +6 -6
  78. package/dist/docs/skills/framework-map.html +5 -5
  79. package/dist/docs/skills/github.html +6 -6
  80. package/dist/docs/skills/hillclimb.html +6 -6
  81. package/dist/docs/skills/index.html +5 -5
  82. package/dist/docs/skills/mcp-auth.html +6 -6
  83. package/dist/docs/skills/otel.html +8 -8
  84. package/dist/docs/skills/otel.md +3 -3
  85. package/dist/docs/skills/setup-slack.html +6 -6
  86. package/dist/docs/storage.html +6 -6
  87. package/dist/docs/templates/agentic-owners.html +6 -6
  88. package/dist/docs/templates/agents-md.html +6 -6
  89. package/dist/docs/templates/code-wiki.html +6 -6
  90. package/dist/docs/templates/demo.html +6 -6
  91. package/dist/docs/templates/grokbot-agents.html +6 -6
  92. package/dist/docs/templates/pr-autofixer.html +6 -6
  93. package/dist/docs/templates/security-help.html +6 -6
  94. package/dist/docs/templates/security-reviewer.html +6 -6
  95. package/dist/docs/templates/triage.html +6 -6
  96. package/dist/docs/troubleshooting.html +5 -5
  97. package/dist/index.d.ts +4 -5
  98. package/dist/index.d.ts.map +1 -1
  99. package/dist/index.js +3 -4
  100. package/dist/internal/cli-deploy.d.ts.map +1 -1
  101. package/dist/internal/cli-deploy.js +23 -11
  102. package/dist/internal/cursor/backend-client.d.ts +25 -0
  103. package/dist/internal/cursor/backend-client.d.ts.map +1 -1
  104. package/dist/internal/cursor/backend-client.js +57 -2
  105. package/dist/internal/deploy-client.d.ts +8 -0
  106. package/dist/internal/deploy-client.d.ts.map +1 -1
  107. package/dist/internal/deploy-client.js +14 -0
  108. package/dist/internal/deploy-manifest.d.ts +3 -0
  109. package/dist/internal/deploy-manifest.d.ts.map +1 -1
  110. package/dist/internal/deploy-manifest.js +15 -12
  111. package/dist/internal/discovery/agent.d.ts.map +1 -1
  112. package/dist/internal/discovery/agent.js +1 -0
  113. package/dist/internal/discovery/info.d.ts.map +1 -1
  114. package/dist/internal/discovery/info.js +7 -3
  115. package/dist/internal/discovery/modules.d.ts +1 -1
  116. package/dist/internal/discovery/modules.d.ts.map +1 -1
  117. package/dist/internal/discovery/modules.js +1 -0
  118. package/dist/internal/discovery/project.d.ts.map +1 -1
  119. package/dist/internal/discovery/project.js +3 -0
  120. package/dist/internal/discovery/reminders.d.ts +7 -0
  121. package/dist/internal/discovery/reminders.d.ts.map +1 -0
  122. package/dist/internal/discovery/reminders.js +65 -0
  123. package/dist/internal/discovery/subagent.js +1 -0
  124. package/dist/internal/distribution.d.ts +5 -0
  125. package/dist/internal/distribution.d.ts.map +1 -1
  126. package/dist/internal/distribution.js +17 -0
  127. package/dist/internal/framework-storage-selection.d.ts +14 -3
  128. package/dist/internal/framework-storage-selection.d.ts.map +1 -1
  129. package/dist/internal/framework-storage-selection.js +36 -6
  130. package/dist/internal/guest-network.d.ts.map +1 -1
  131. package/dist/internal/guest-network.js +2 -0
  132. package/dist/internal/hosted-delivery-protocol.d.ts +48 -0
  133. package/dist/internal/hosted-delivery-protocol.d.ts.map +1 -1
  134. package/dist/internal/hosted-delivery-protocol.js +60 -0
  135. package/dist/internal/hosted-delivery.d.ts +10 -0
  136. package/dist/internal/hosted-delivery.d.ts.map +1 -1
  137. package/dist/internal/hosted-delivery.js +83 -18
  138. package/dist/internal/local-control-plane.d.ts +18 -2
  139. package/dist/internal/local-control-plane.d.ts.map +1 -1
  140. package/dist/internal/local-control-plane.js +199 -17
  141. package/dist/internal/reminder-control-plane-protocol.d.ts +187 -0
  142. package/dist/internal/reminder-control-plane-protocol.d.ts.map +1 -0
  143. package/dist/internal/reminder-control-plane-protocol.js +244 -0
  144. package/dist/internal/reminder-runner.d.ts +48 -6
  145. package/dist/internal/reminder-runner.d.ts.map +1 -1
  146. package/dist/internal/reminder-runner.js +336 -48
  147. package/dist/internal/reminder-schedule.d.ts +5 -2
  148. package/dist/internal/reminder-schedule.d.ts.map +1 -1
  149. package/dist/internal/reminder-schedule.js +8 -6
  150. package/dist/internal/reminder-store.d.ts +24 -7
  151. package/dist/internal/reminder-store.d.ts.map +1 -1
  152. package/dist/internal/reminder-store.js +6 -5
  153. package/dist/internal/request-error-status.d.ts +2 -0
  154. package/dist/internal/request-error-status.d.ts.map +1 -1
  155. package/dist/internal/request-error-status.js +22 -3
  156. package/dist/internal/scm/pr-url.d.ts +40 -0
  157. package/dist/internal/scm/pr-url.d.ts.map +1 -1
  158. package/dist/internal/scm/pr-url.js +93 -14
  159. package/dist/internal/server.d.ts.map +1 -1
  160. package/dist/internal/server.js +67 -27
  161. package/dist/internal/session-engine.d.ts +4 -0
  162. package/dist/internal/session-engine.d.ts.map +1 -1
  163. package/dist/internal/session-engine.js +5 -0
  164. package/dist/playground/assets/{index-BaQzjaNK.js → index-Bq2HpEQB.js} +23 -23
  165. package/dist/playground/assets/index-CZKKNlmb.css +1 -0
  166. package/dist/playground/index.html +2 -2
  167. package/dist/reminders.d.ts +24 -5
  168. package/dist/reminders.d.ts.map +1 -1
  169. package/dist/reminders.js +28 -5
  170. package/dist/types.d.ts +89 -20
  171. package/dist/types.d.ts.map +1 -1
  172. package/docs/guides/opentelemetry.md +3 -4
  173. package/docs/reference/channels.md +11 -19
  174. package/docs/reference/cli.md +5 -2
  175. package/package.json +3 -3
  176. package/skills/otel/SKILL.md +3 -3
  177. package/src/bin/agent-serve.ts +1 -9
  178. package/src/channels/bitbucket/types.ts +2 -4
  179. package/src/channels/deployments/types.ts +5 -8
  180. package/src/channels/github/types.ts +2 -4
  181. package/src/channels/gitlab/types.ts +2 -5
  182. package/src/channels/origin/types.ts +2 -3
  183. package/src/channels/slack/defaults.ts +3 -0
  184. package/src/channels/slack/dispatch.ts +1 -0
  185. package/src/channels/slack/live-delivery.ts +53 -12
  186. package/src/channels/slack/types.ts +4 -3
  187. package/src/index.ts +6 -3
  188. package/src/internal/cli-deploy.ts +30 -12
  189. package/src/internal/cursor/backend-client.ts +90 -2
  190. package/src/internal/deploy-client.ts +16 -0
  191. package/src/internal/deploy-manifest.ts +16 -11
  192. package/src/internal/discovery/agent.ts +1 -0
  193. package/src/internal/discovery/info.ts +6 -0
  194. package/src/internal/discovery/modules.ts +2 -4
  195. package/src/internal/discovery/project.ts +3 -0
  196. package/src/internal/discovery/reminders.ts +65 -0
  197. package/src/internal/discovery/subagent.ts +1 -0
  198. package/src/internal/distribution.ts +20 -0
  199. package/src/internal/framework-storage-selection.ts +43 -6
  200. package/src/internal/guest-network.ts +3 -0
  201. package/src/internal/hosted-delivery-protocol.ts +96 -0
  202. package/src/internal/hosted-delivery.ts +154 -27
  203. package/src/internal/local-control-plane.ts +281 -22
  204. package/src/internal/reminder-control-plane-protocol.ts +499 -0
  205. package/src/internal/reminder-runner.ts +421 -48
  206. package/src/internal/reminder-schedule.ts +11 -6
  207. package/src/internal/reminder-store.ts +27 -7
  208. package/src/internal/request-error-status.ts +27 -3
  209. package/src/internal/scm/pr-url.ts +167 -20
  210. package/src/internal/server.ts +69 -16
  211. package/src/internal/session-engine.ts +5 -0
  212. package/src/reminders.ts +35 -4
  213. package/src/types.ts +95 -20
  214. package/dist/docs/assets/chunks/@localSearchIndexroot.CJxMklAy.js +0 -1
  215. package/dist/playground/assets/index-CceqnLTX.css +0 -1
  216. /package/dist/docs/assets/{reference_cli.md.C6Lo8SEd.lean.js → reference_cli.md.Dm67hd2D.lean.js} +0 -0
  217. /package/dist/docs/assets/{skills_otel.md.CqhobH3B.lean.js → skills_otel.md.B5Ddv1xH.lean.js} +0 -0
@@ -6,12 +6,12 @@
6
6
  <title>Evals | Agent SDK</title>
7
7
  <meta name="description" content="Define repeatable checks with defineEval, run them with agent-sdk eval, and use them as regression checks.">
8
8
  <meta name="generator" content="VitePress v1.6.4">
9
- <link rel="preload stylesheet" href="/docs/assets/style.BRuM8477.css" as="style">
9
+ <link rel="preload stylesheet" href="/docs/assets/style.DCD_zyg3.css" as="style">
10
10
  <link rel="preload stylesheet" href="/docs/vp-icons.css" as="style">
11
11
 
12
- <script type="module" src="/docs/assets/app.BVwaffdK.js"></script>
12
+ <script type="module" src="/docs/assets/app.CtuckIiL.js"></script>
13
13
  <link rel="preload" href="/docs/assets/inter-roman-latin.Di8DUHzh.woff2" as="font" type="font/woff2" crossorigin="">
14
- <link rel="modulepreload" href="/docs/assets/chunks/theme.BYdI_6sF.js">
14
+ <link rel="modulepreload" href="/docs/assets/chunks/theme.DhpIe0Pa.js">
15
15
  <link rel="modulepreload" href="/docs/assets/chunks/framework.BCISBCiQ.js">
16
16
  <link rel="modulepreload" href="/docs/assets/evals.md.CbMoebP1.lean.js">
17
17
  <script id="check-dark-mode">(()=>{const e=localStorage.getItem("vitepress-theme-appearance")||"auto",a=window.matchMedia("(prefers-color-scheme: dark)").matches;(!e||e==="auto"?a:e==="dark")&&document.documentElement.classList.add("dark")})();</script>
@@ -20,7 +20,7 @@
20
20
  <link rel="alternate" type="text/markdown" href="/docs/evals.md">
21
21
  </head>
22
22
  <body>
23
- <div id="app"><div class="Layout" data-v-282c430e><!--[--><!--]--><!--[--><span tabindex="-1" data-v-af8643bc></span><a href="#VPContent" class="VPSkipLink visually-hidden" data-v-af8643bc>Skip to content</a><!--]--><!----><header class="VPNav" data-v-282c430e data-v-4751689a><div class="VPNavBar" data-v-4751689a data-v-d2a336f3><div class="wrapper" data-v-d2a336f3><div class="container" data-v-d2a336f3><div class="title" data-v-d2a336f3><div class="VPNavBarTitle has-sidebar" data-v-d2a336f3 data-v-3a787a7b><a class="title" href="/docs/" data-v-3a787a7b><!--[--><!--]--><!----><span data-v-3a787a7b>Agent SDK</span><!--[--><!--[--><!--[--><!--[--><span class="agent-sdk-version" title="@cursor/july 0.1.105" data-v-c26a5f4f>0.1.105</span><!--]--><!--]--><!--]--><!--]--></a></div></div><div class="content" data-v-d2a336f3><div class="content-body" data-v-d2a336f3><!--[--><!--]--><div class="VPNavBarSearch search" data-v-d2a336f3><!--[--><!----><div id="local-search"><button type="button" class="DocSearch DocSearch-Button" aria-label="Search"><span class="DocSearch-Button-Container"><span class="vp-icon DocSearch-Search-Icon"></span><span class="DocSearch-Button-Placeholder">Search</span></span><span class="DocSearch-Button-Keys"><kbd class="DocSearch-Button-Key"></kbd><kbd class="DocSearch-Button-Key">K</kbd></span></button></div><!--]--></div><nav aria-labelledby="main-nav-aria-label" class="VPNavBarMenu menu" data-v-d2a336f3 data-v-b60c0a58><span id="main-nav-aria-label" class="visually-hidden" data-v-b60c0a58> Main Navigation </span><!--[--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/quickstart.html" tabindex="0" data-v-b60c0a58 data-v-74de87b9><!--[--><span data-v-74de87b9>Quickstart</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/guides/webhooks.html" tabindex="0" data-v-b60c0a58 data-v-74de87b9><!--[--><span data-v-74de87b9>Guides</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/templates/demo.html" tabindex="0" data-v-b60c0a58 data-v-74de87b9><!--[--><span data-v-74de87b9>Templates</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/reference/project-layout.html" tabindex="0" data-v-b60c0a58 data-v-74de87b9><!--[--><span data-v-74de87b9>Reference</span><!--]--></a><!--]--><!--]--></nav><!----><div class="VPNavBarAppearance appearance" data-v-d2a336f3 data-v-d559fc20><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-d559fc20 data-v-deaafa36 data-v-88da49d2><span class="check" data-v-88da49d2><span class="icon" data-v-88da49d2><!--[--><span class="vpi-sun sun" data-v-deaafa36></span><span class="vpi-moon moon" data-v-deaafa36></span><!--]--></span></span></button></div><!----><div class="VPFlyout VPNavBarExtra extra" data-v-d2a336f3 data-v-75b97eb1 data-v-ed271fd4><button type="button" class="button" aria-haspopup="true" aria-expanded="false" aria-label="extra navigation" data-v-ed271fd4><span class="vpi-more-horizontal icon" data-v-ed271fd4></span></button><div class="menu" data-v-ed271fd4><div class="VPMenu" data-v-ed271fd4 data-v-505057d2><!----><!--[--><!--[--><!----><div class="group" data-v-75b97eb1><div class="item appearance" data-v-75b97eb1><p class="label" data-v-75b97eb1>Appearance</p><div class="appearance-action" data-v-75b97eb1><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-75b97eb1 data-v-deaafa36 data-v-88da49d2><span class="check" data-v-88da49d2><span class="icon" data-v-88da49d2><!--[--><span class="vpi-sun sun" data-v-deaafa36></span><span class="vpi-moon moon" data-v-deaafa36></span><!--]--></span></span></button></div></div></div><!----><!--]--><!--]--></div></div></div><!--[--><!--]--><button type="button" class="VPNavBarHamburger hamburger" aria-label="mobile navigation" aria-expanded="false" aria-controls="VPNavScreen" data-v-d2a336f3 data-v-966f1ac7><span class="container" data-v-966f1ac7><span class="top" data-v-966f1ac7></span><span class="middle" data-v-966f1ac7></span><span class="bottom" data-v-966f1ac7></span></span></button></div></div></div></div><div class="divider" data-v-d2a336f3><div class="divider-line" data-v-d2a336f3></div></div></div><!----></header><div class="VPLocalNav has-sidebar empty" data-v-282c430e data-v-f96f8409><div class="container" data-v-f96f8409><button class="menu" aria-expanded="false" aria-controls="VPSidebarNav" data-v-f96f8409><span class="vpi-align-left menu-icon" data-v-f96f8409></span><span class="menu-text" data-v-f96f8409>Menu</span></button><div class="VPLocalNavOutlineDropdown" style="--vp-vh:0px;" data-v-f96f8409 data-v-1bc67f4b><button data-v-1bc67f4b>Return to top</button><!----></div></div></div><aside class="VPSidebar" data-v-282c430e data-v-3d1258d2><div class="curtain" data-v-3d1258d2></div><nav class="nav" id="VPSidebarNav" aria-labelledby="sidebar-aria-label" tabindex="-1" data-v-3d1258d2><span class="visually-hidden" id="sidebar-aria-label" data-v-3d1258d2> Sidebar Navigation </span><!--[--><!--]--><!--[--><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Getting started</h2><!----></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/quickstart.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Quickstart</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/scaffolding-agents.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Scaffold an agent with Cursor</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/convert-automation.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Convert a Cursor Automation</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/concepts.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Concepts</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Templates</h2><!----></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/demo.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Demo agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/grokbot-agents.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Grok Bot agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/code-wiki.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Code wiki</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/agents-md.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Living AGENTS.md</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/security-reviewer.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Security reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/security-help.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Security help</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/agentic-owners.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Agentic Owners</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/pr-autofixer.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>PR autofixer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/templates/triage.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Triage agent</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0 has-active" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Improving agents</h2><!----></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/building-with-agents.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Building agents with agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/evals.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/ab.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Live A/B metrics</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/storage.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Storage</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/hillclimbing.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Hillclimbing</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0 collapsible collapsed" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Coding-agent skills</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-61fcdc7c><span class="vpi-chevron-right caret-icon" data-v-61fcdc7c></span></div></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/ab.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>ab</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/create-agent.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>create-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/debug.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>debug</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/evals.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/framework-map.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>framework-map</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/github.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>github</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/hillclimb.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>hillclimb</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/mcp-auth.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>mcp-auth</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/otel.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>otel</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/skills/setup-slack.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>setup-slack</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Guides</h2><!----></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/webhooks.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Webhooks & custom channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/github.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>GitHub</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/slack.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Slack</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/human-in-the-loop.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Human-in-the-loop approvals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/agent-to-agent.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Agent-to-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/mcp-oauth.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Host MCP OAuth</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/cloud-agents.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Cursor cloud agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/grokbot-agents.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Cursor Grok Bot agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/guides/opentelemetry.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>OpenTelemetry</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Operating</h2><!----></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/deployment.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Deployment</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/troubleshooting.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Fix common problems</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-f2306e18><section class="VPSidebarItem level-0 collapsible" data-v-f2306e18 data-v-61fcdc7c><div class="item" role="button" tabindex="0" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><h2 class="text" data-v-61fcdc7c>Reference</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-61fcdc7c><span class="vpi-chevron-right caret-icon" data-v-61fcdc7c></span></div></div><div class="items" data-v-61fcdc7c><!--[--><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/project-layout.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Project layout</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/agent-config.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Agent config</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/instructions.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Instructions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/tools.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Tools</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/prompt.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>prompt</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/skills.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Skills</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/connections.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>MCP Connections</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/subagents.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Subagents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/extensions.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Extensions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/channels.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/schedules.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Schedules & reminders</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/hooks.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Hooks</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/artifacts.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Artifacts</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/result.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Turn result</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/sessions.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Sessions & streaming</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/playground.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>Playground</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/cli.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>CLI</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-61fcdc7c><div class="item" data-v-61fcdc7c><div class="indicator" data-v-61fcdc7c></div><a class="VPLink link link" href="/docs/reference/http-api.html" data-v-61fcdc7c><!--[--><p class="text" data-v-61fcdc7c>HTTP API</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><!--]--><!--[--><!--]--></nav></aside><div class="VPContent has-sidebar" id="VPContent" data-v-282c430e data-v-3b8d81ee><div class="VPDoc has-sidebar has-aside" data-v-3b8d81ee data-v-2e506a34><!--[--><!--]--><div class="container" data-v-2e506a34><div class="aside" data-v-2e506a34><div class="aside-curtain" data-v-2e506a34></div><div class="aside-container" data-v-2e506a34><div class="aside-content" data-v-2e506a34><div class="VPDocAside" data-v-2e506a34 data-v-b485a79e><!--[--><!--]--><!--[--><!--]--><nav aria-labelledby="doc-outline-aria-label" class="VPDocAsideOutline" data-v-b485a79e data-v-cc6b345c><div class="content" data-v-cc6b345c><div class="outline-marker" data-v-cc6b345c></div><div aria-level="2" class="outline-title" id="doc-outline-aria-label" role="heading" data-v-cc6b345c>On this page</div><ul class="VPDocOutlineItem root" data-v-cc6b345c data-v-4f0215bf><!--[--><!--]--></ul></div></nav><!--[--><!--]--><div class="spacer" data-v-b485a79e></div><!--[--><!--]--><!----><!--[--><!--]--><!--[--><!--]--></div></div></div></div><div class="content" data-v-2e506a34><div class="content-container" data-v-2e506a34><!--[--><!--]--><main class="main" data-v-2e506a34><div style="position:relative;" class="vp-doc _docs_evals" data-v-2e506a34><div><h1 id="evals" tabindex="-1">Evals <a class="header-anchor" href="#evals" aria-label="Permalink to &quot;Evals&quot;">​</a></h1><p>An eval is a repeatable check that runs your agent against a fixed input and gates the recorded trajectory: the run completed, the right tool ran, the reply has the right shape. Evals are how you know a prompt tweak helped, a refactor didn&#39;t regress the agent, and last month&#39;s fix is still holding.</p><p>Evals exercise the same surface your users hit. The runner starts (or targets) a real agent server, drives sessions over the public API, and grades what comes back. A passing eval means the agent started, accepted a message, and did what you asserted.</p><h2 id="define-evals-with-defineeval" tabindex="-1">Define evals with <code>defineEval</code> <a class="header-anchor" href="#define-evals-with-defineeval" aria-label="Permalink to &quot;Define evals with `defineEval`&quot;">​</a></h2><p>The Agent SDK discovers evals under the project-root <code>evals/</code> directory, in <code>.eval.ts</code> or <code>.eval.js</code> files. That&#39;s a sibling of <code>agent/</code>, never inside it (<code>agent/evals/</code> is silently ignored). TypeScript is the normal authoring format.</p><p>The file path is the eval&#39;s identity, so you don&#39;t author an id. Directories group related evals: <code>evals/builds/api.eval.ts</code> becomes id <code>builds/api</code>. An <code>index</code> filename collapses to its directory, so <code>evals/builds/index.eval.ts</code> becomes <code>builds</code>.</p><p>An eval is a single <code>async test(t)</code>. You drive the agent with <code>t</code> and assert on the run with the same <code>t</code>:</p><div class="language-ts vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">ts</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">// evals/readiness.eval.ts</span></span>
23
+ <div id="app"><div class="Layout" data-v-990b7706><!--[--><!--]--><!--[--><span tabindex="-1" data-v-0b74bec0></span><a href="#VPContent" class="VPSkipLink visually-hidden" data-v-0b74bec0>Skip to content</a><!--]--><!----><header class="VPNav" data-v-990b7706 data-v-af996e33><div class="VPNavBar" data-v-af996e33 data-v-c7f322eb><div class="wrapper" data-v-c7f322eb><div class="container" data-v-c7f322eb><div class="title" data-v-c7f322eb><div class="VPNavBarTitle has-sidebar" data-v-c7f322eb data-v-d1664c59><a class="title" href="/docs/" data-v-d1664c59><!--[--><!--]--><!----><span data-v-d1664c59>Agent SDK</span><!--[--><!--[--><!--[--><!--[--><span class="agent-sdk-version" title="@cursor/july 0.1.107" data-v-c26a5f4f>0.1.107</span><!--]--><!--]--><!--]--><!--]--></a></div></div><div class="content" data-v-c7f322eb><div class="content-body" data-v-c7f322eb><!--[--><!--]--><div class="VPNavBarSearch search" data-v-c7f322eb><!--[--><!----><div id="local-search"><button type="button" class="DocSearch DocSearch-Button" aria-label="Search"><span class="DocSearch-Button-Container"><span class="vp-icon DocSearch-Search-Icon"></span><span class="DocSearch-Button-Placeholder">Search</span></span><span class="DocSearch-Button-Keys"><kbd class="DocSearch-Button-Key"></kbd><kbd class="DocSearch-Button-Key">K</kbd></span></button></div><!--]--></div><nav aria-labelledby="main-nav-aria-label" class="VPNavBarMenu menu" data-v-c7f322eb data-v-ab749b05><span id="main-nav-aria-label" class="visually-hidden" data-v-ab749b05> Main Navigation </span><!--[--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/quickstart.html" tabindex="0" data-v-ab749b05 data-v-ef8d136f><!--[--><span data-v-ef8d136f>Quickstart</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/guides/webhooks.html" tabindex="0" data-v-ab749b05 data-v-ef8d136f><!--[--><span data-v-ef8d136f>Guides</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/templates/demo.html" tabindex="0" data-v-ab749b05 data-v-ef8d136f><!--[--><span data-v-ef8d136f>Templates</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/reference/project-layout.html" tabindex="0" data-v-ab749b05 data-v-ef8d136f><!--[--><span data-v-ef8d136f>Reference</span><!--]--></a><!--]--><!--]--></nav><!----><div class="VPNavBarAppearance appearance" data-v-c7f322eb data-v-7c6c0ede><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-7c6c0ede data-v-87b8a6bb data-v-9f798512><span class="check" data-v-9f798512><span class="icon" data-v-9f798512><!--[--><span class="vpi-sun sun" data-v-87b8a6bb></span><span class="vpi-moon moon" data-v-87b8a6bb></span><!--]--></span></span></button></div><!----><div class="VPFlyout VPNavBarExtra extra" data-v-c7f322eb data-v-e53d4296 data-v-79fc383b><button type="button" class="button" aria-haspopup="true" aria-expanded="false" aria-label="extra navigation" data-v-79fc383b><span class="vpi-more-horizontal icon" data-v-79fc383b></span></button><div class="menu" data-v-79fc383b><div class="VPMenu" data-v-79fc383b data-v-0d780c53><!----><!--[--><!--[--><!----><div class="group" data-v-e53d4296><div class="item appearance" data-v-e53d4296><p class="label" data-v-e53d4296>Appearance</p><div class="appearance-action" data-v-e53d4296><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-e53d4296 data-v-87b8a6bb data-v-9f798512><span class="check" data-v-9f798512><span class="icon" data-v-9f798512><!--[--><span class="vpi-sun sun" data-v-87b8a6bb></span><span class="vpi-moon moon" data-v-87b8a6bb></span><!--]--></span></span></button></div></div></div><!----><!--]--><!--]--></div></div></div><!--[--><!--]--><button type="button" class="VPNavBarHamburger hamburger" aria-label="mobile navigation" aria-expanded="false" aria-controls="VPNavScreen" data-v-c7f322eb data-v-edf597df><span class="container" data-v-edf597df><span class="top" data-v-edf597df></span><span class="middle" data-v-edf597df></span><span class="bottom" data-v-edf597df></span></span></button></div></div></div></div><div class="divider" data-v-c7f322eb><div class="divider-line" data-v-c7f322eb></div></div></div><!----></header><div class="VPLocalNav has-sidebar empty" data-v-990b7706 data-v-054265d6><div class="container" data-v-054265d6><button class="menu" aria-expanded="false" aria-controls="VPSidebarNav" data-v-054265d6><span class="vpi-align-left menu-icon" data-v-054265d6></span><span class="menu-text" data-v-054265d6>Menu</span></button><div class="VPLocalNavOutlineDropdown" style="--vp-vh:0px;" data-v-054265d6 data-v-65e16bd4><button data-v-65e16bd4>Return to top</button><!----></div></div></div><aside class="VPSidebar" data-v-990b7706 data-v-03069692><div class="curtain" data-v-03069692></div><nav class="nav" id="VPSidebarNav" aria-labelledby="sidebar-aria-label" tabindex="-1" data-v-03069692><span class="visually-hidden" id="sidebar-aria-label" data-v-03069692> Sidebar Navigation </span><!--[--><!--]--><!--[--><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Getting started</h2><!----></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/quickstart.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Quickstart</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/scaffolding-agents.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Scaffold an agent with Cursor</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/convert-automation.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Convert a Cursor Automation</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/concepts.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Concepts</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Templates</h2><!----></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/demo.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Demo agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/grokbot-agents.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Grok Bot agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/code-wiki.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Code wiki</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/agents-md.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Living AGENTS.md</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/security-reviewer.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Security reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/security-help.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Security help</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/agentic-owners.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Agentic Owners</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/pr-autofixer.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>PR autofixer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/templates/triage.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Triage agent</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0 has-active" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Improving agents</h2><!----></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/building-with-agents.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Building agents with agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/evals.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/ab.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Live A/B metrics</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/storage.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Storage</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/hillclimbing.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Hillclimbing</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0 collapsible collapsed" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Coding-agent skills</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-9a4aac52><span class="vpi-chevron-right caret-icon" data-v-9a4aac52></span></div></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/ab.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>ab</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/create-agent.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>create-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/debug.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>debug</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/evals.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/framework-map.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>framework-map</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/github.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>github</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/hillclimb.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>hillclimb</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/mcp-auth.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>mcp-auth</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/otel.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>otel</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/skills/setup-slack.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>setup-slack</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Guides</h2><!----></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/webhooks.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Webhooks & custom channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/github.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>GitHub</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/slack.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Slack</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/human-in-the-loop.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Human-in-the-loop approvals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/agent-to-agent.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Agent-to-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/mcp-oauth.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Host MCP OAuth</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/cloud-agents.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Cursor cloud agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/grokbot-agents.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Cursor Grok Bot agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/guides/opentelemetry.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>OpenTelemetry</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Operating</h2><!----></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/deployment.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Deployment</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/troubleshooting.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Fix common problems</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-53da167f><section class="VPSidebarItem level-0 collapsible" data-v-53da167f data-v-9a4aac52><div class="item" role="button" tabindex="0" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><h2 class="text" data-v-9a4aac52>Reference</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-9a4aac52><span class="vpi-chevron-right caret-icon" data-v-9a4aac52></span></div></div><div class="items" data-v-9a4aac52><!--[--><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/project-layout.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Project layout</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/agent-config.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Agent config</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/instructions.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Instructions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/tools.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Tools</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/prompt.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>prompt</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/skills.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Skills</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/connections.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>MCP Connections</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/subagents.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Subagents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/extensions.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Extensions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/channels.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/schedules.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Schedules & reminders</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/hooks.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Hooks</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/artifacts.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Artifacts</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/result.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Turn result</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/sessions.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Sessions & streaming</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/playground.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>Playground</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/cli.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>CLI</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-9a4aac52><div class="item" data-v-9a4aac52><div class="indicator" data-v-9a4aac52></div><a class="VPLink link link" href="/docs/reference/http-api.html" data-v-9a4aac52><!--[--><p class="text" data-v-9a4aac52>HTTP API</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><!--]--><!--[--><!--]--></nav></aside><div class="VPContent has-sidebar" id="VPContent" data-v-990b7706 data-v-7a16121a><div class="VPDoc has-sidebar has-aside" data-v-7a16121a data-v-38549ef7><!--[--><!--]--><div class="container" data-v-38549ef7><div class="aside" data-v-38549ef7><div class="aside-curtain" data-v-38549ef7></div><div class="aside-container" data-v-38549ef7><div class="aside-content" data-v-38549ef7><div class="VPDocAside" data-v-38549ef7 data-v-da90486d><!--[--><!--]--><!--[--><!--]--><nav aria-labelledby="doc-outline-aria-label" class="VPDocAsideOutline" data-v-da90486d data-v-0895f2d7><div class="content" data-v-0895f2d7><div class="outline-marker" data-v-0895f2d7></div><div aria-level="2" class="outline-title" id="doc-outline-aria-label" role="heading" data-v-0895f2d7>On this page</div><ul class="VPDocOutlineItem root" data-v-0895f2d7 data-v-0cae2091><!--[--><!--]--></ul></div></nav><!--[--><!--]--><div class="spacer" data-v-da90486d></div><!--[--><!--]--><!----><!--[--><!--]--><!--[--><!--]--></div></div></div></div><div class="content" data-v-38549ef7><div class="content-container" data-v-38549ef7><!--[--><!--]--><main class="main" data-v-38549ef7><div style="position:relative;" class="vp-doc _docs_evals" data-v-38549ef7><div><h1 id="evals" tabindex="-1">Evals <a class="header-anchor" href="#evals" aria-label="Permalink to &quot;Evals&quot;">​</a></h1><p>An eval is a repeatable check that runs your agent against a fixed input and gates the recorded trajectory: the run completed, the right tool ran, the reply has the right shape. Evals are how you know a prompt tweak helped, a refactor didn&#39;t regress the agent, and last month&#39;s fix is still holding.</p><p>Evals exercise the same surface your users hit. The runner starts (or targets) a real agent server, drives sessions over the public API, and grades what comes back. A passing eval means the agent started, accepted a message, and did what you asserted.</p><h2 id="define-evals-with-defineeval" tabindex="-1">Define evals with <code>defineEval</code> <a class="header-anchor" href="#define-evals-with-defineeval" aria-label="Permalink to &quot;Define evals with `defineEval`&quot;">​</a></h2><p>The Agent SDK discovers evals under the project-root <code>evals/</code> directory, in <code>.eval.ts</code> or <code>.eval.js</code> files. That&#39;s a sibling of <code>agent/</code>, never inside it (<code>agent/evals/</code> is silently ignored). TypeScript is the normal authoring format.</p><p>The file path is the eval&#39;s identity, so you don&#39;t author an id. Directories group related evals: <code>evals/builds/api.eval.ts</code> becomes id <code>builds/api</code>. An <code>index</code> filename collapses to its directory, so <code>evals/builds/index.eval.ts</code> becomes <code>builds</code>.</p><p>An eval is a single <code>async test(t)</code>. You drive the agent with <code>t</code> and assert on the run with the same <code>t</code>:</p><div class="language-ts vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">ts</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">// evals/readiness.eval.ts</span></span>
24
24
  <span class="line"><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">import</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> { defineEval, includes } </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">from</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &quot;@cursor/july/evals&quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">;</span></span>
25
25
  <span class="line"></span>
26
26
  <span class="line"><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">export</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;"> default</span><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;"> defineEval</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">({</span></span>
@@ -104,8 +104,8 @@
104
104
  <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># Playground: https://…/playground?view=evals&amp;evalRunId=evalrun_…</span></span>
105
105
  <span class="line"></span>
106
106
  <span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agent-sdk</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> cancel</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> evalrun_…</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --prod</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --slug</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> vulnerability-scanner</span></span>
107
- <span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agent-sdk</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> status</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> evalrun_…</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --prod</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --slug</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> vulnerability-scanner</span></span></code></pre></div><h2 id="what-good-cases-assert" tabindex="-1">What good cases assert <a class="header-anchor" href="#what-good-cases-assert" aria-label="Permalink to &quot;What good cases assert&quot;">​</a></h2><p>Gate decisions and shape, not prose. Model wording varies run to run. Tool choice, tool avoidance, and output structure are the stable contract.</p><ol><li><code>t.succeeded()</code>: always, first.</li><li>The tool decision: <code>calledTool</code> for the intended path, <code>notCalledTool</code> for the likely wrong alternative. The pair is stronger than either alone.</li><li>Output shape: a regex for the contract (<code>/ready|blocked/i</code>, a JSON marker, a findings-block fence), never exact sentences.</li><li>For structured output, parse <code>t.reply</code> and check fields with <code>satisfies</code> instead of substring-matching JSON.</li></ol><p>The common failure modes: asserting exact phrasing, packing more than about five gates into one case (split it), and cases that depend on live external state that drifts (pin the input; see fixtures).</p><h2 id="pick-fixtures-by-agent-type" tabindex="-1">Pick fixtures by agent type <a class="header-anchor" href="#pick-fixtures-by-agent-type" aria-label="Permalink to &quot;Pick fixtures by agent type&quot;">​</a></h2><p>The right fixture depends on the surface under test.</p><table tabindex="0"><thead><tr><th>Agent surface</th><th>Fixture</th></tr></thead><tbody><tr><td>Chat / domain assistant</td><td>A canonical prompt string, chosen once and frozen</td></tr><tr><td>Tool-heavy</td><td>Run <code>agent-sdk call &lt;tool&gt;</code> first to pin what the tool returns, then freeze the prompt that triggers it</td></tr><tr><td>GitHub webhook</td><td><code>agent-sdk github replay &lt;pr&gt; --events &#39;*&#39; --dry-run --out fixtures/github</code> snapshots real payloads for offline replay (<a href="./guides/github.html">GitHub guide</a>)</td></tr><tr><td>PR reviewer with host preparation</td><td>Diff, metadata, and gold labels pinned to commit SHAs; keep any live PR matrix small</td></tr><tr><td>Workspace-dependent</td><td><code>workspaceFiles</code> in <code>t.send</code> options, never developer-machine paths</td></tr></tbody></table><p>Tag the fast, reliably passing core <code>smoke</code> and run <code>--tag smoke</code> in the inner loop. Leave slow or flaky-prone cases untagged for explicit runs.</p><h3 id="materialize-api-backed-fixtures" tabindex="-1">Materialize API-backed fixtures <a class="header-anchor" href="#materialize-api-backed-fixtures" aria-label="Permalink to &quot;Materialize API-backed fixtures&quot;">​</a></h3><p>An input that only points at external data, such as a pull request URL, snapshot id, or pair of commit SHAs, is not self-contained. Fetch it once and commit the rendered fixture before you expand the suite.</p><ol><li>Save the diff, metadata, and labels under <code>fixtures/</code> at pinned revisions.</li><li>Seed those files with <code>workspaceFiles</code>, or read them from the fixture directory.</li><li>Assert decisions and output shape against the saved evidence.</li><li>Keep a small <code>smoke</code> subset for any remaining live pipeline checks.</li></ol><p>Read committed fixtures with <code>@cursor/july/evals/loaders</code>: <code>loadJson</code>, <code>loadJsonl</code>, and <code>loadYaml</code> resolve relative paths against the project root the runner discovered, not the cwd the CLI was invoked from (<code>resolveFixturePath</code> and <code>evalFixtureRoot</code> expose the same resolution for other file formats).</p><p><code>maxConcurrency</code> limits parallel datapoints. It does not limit model or API fan-out inside one datapoint. Materialized fixtures prevent a large suite from exhausting provider and GitHub rate limits. The <a href="./skills/evals.html">evals skill</a> has the full fixture workflow.</p><h2 id="keep-improvements-with-regression-evals" tabindex="-1">Keep improvements with regression evals <a class="header-anchor" href="#keep-improvements-with-regression-evals" aria-label="Permalink to &quot;Keep improvements with regression evals&quot;">​</a></h2><p>Every <a href="./hillclimbing.html">hillclimb</a> round that keeps a change must land an eval that would have failed before the change. If you can&#39;t express the improvement as a gate (a <code>calledTool</code> shift, a bounded <code>action.result</code> count, an output-shape regex), the improvement is unverified, and it&#39;ll regress silently.</p><p>The rule cuts the other way too: never weaken an existing gate to make a round pass. That&#39;s the freeze line moving, and it turns your regression suite into a list of checks that no longer protect anything.</p><h2 id="compare-variants-on-live-traffic" tabindex="-1">Compare variants on live traffic <a class="header-anchor" href="#compare-variants-on-live-traffic" aria-label="Permalink to &quot;Compare variants on live traffic&quot;">​</a></h2><p>Use <code>defineAB</code> to compare variant metrics on live sessions. It is not a test runner and has no <code>agent-sdk ab</code> command. Keep <code>defineEval</code> as the regression ratchet. Eval sessions do not enroll or change live metrics. See <a href="./ab.html">Live A/B metrics</a> for assignment, behavior, collection, and inspection.</p><h2 id="what-s-next" tabindex="-1">What&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./ab.html">Live A/B metrics</a>: sticky variants and cumulative metrics on live sessions</li><li><a href="./hillclimbing.html">Hillclimbing</a>: the loop evals make trustworthy</li><li><a href="./building-with-agents.html">Building agents with agents</a>: have a coding agent write the first suite</li><li><a href="./guides/github.html">GitHub guide</a>: deterministic webhook fixtures with <code>github replay</code></li><li><a href="./reference/sessions.html">Sessions and streaming</a>: the events <code>t.events</code> contains</li></ul></div></div></main><footer class="VPDocFooter" data-v-2e506a34 data-v-574f06fa><!--[--><!--]--><!----><nav class="prev-next" aria-labelledby="doc-footer-aria-label" data-v-574f06fa><span class="visually-hidden" id="doc-footer-aria-label" data-v-574f06fa>Pager</span><div class="pager" data-v-574f06fa><a class="VPLink link pager-link prev" href="/docs/building-with-agents.html" data-v-574f06fa><!--[--><span class="desc" data-v-574f06fa>Previous page</span><span class="title" data-v-574f06fa>Building agents with agents</span><!--]--></a></div><div class="pager" data-v-574f06fa><a class="VPLink link pager-link next" href="/docs/ab.html" data-v-574f06fa><!--[--><span class="desc" data-v-574f06fa>Next page</span><span class="title" data-v-574f06fa>Live A/B metrics</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
108
- <script>window.__VP_HASH_MAP__=JSON.parse("{\"ab.md\":\"mlVgqvSk\",\"building-with-agents.md\":\"CUSWxlP_\",\"concepts.md\":\"DgEcZOfT\",\"deployment.md\":\"MFPKqsqI\",\"evals.md\":\"CbMoebP1\",\"guides_agent-to-agent.md\":\"BDb0t1QV\",\"guides_cloud-agents.md\":\"DS8RIjwx\",\"guides_convert-automation.md\":\"B4sjlodG\",\"guides_github.md\":\"c0gdGJ-D\",\"guides_grokbot-agents.md\":\"DVqdhBKb\",\"guides_human-in-the-loop.md\":\"By1G2T3_\",\"guides_mcp-oauth.md\":\"Dp6cDP7f\",\"guides_opentelemetry.md\":\"COh9jFVg\",\"guides_slack.md\":\"9oHPye9o\",\"guides_webhooks.md\":\"DKdA43Qm\",\"hillclimbing.md\":\"CpTGTCle\",\"index.md\":\"BW_6tOgR\",\"quickstart.md\":\"DdQOF7Y8\",\"reference_agent-config.md\":\"CHNpiyp4\",\"reference_artifacts.md\":\"Dior32Qw\",\"reference_channels.md\":\"DBGcaB7m\",\"reference_cli.md\":\"C6Lo8SEd\",\"reference_connections.md\":\"Di6jJAXF\",\"reference_extensions.md\":\"CGmMLblt\",\"reference_hooks.md\":\"Ddt5DdgJ\",\"reference_http-api.md\":\"oySXBO8o\",\"reference_instructions.md\":\"CR7XSsGk\",\"reference_playground.md\":\"4myJPxrf\",\"reference_project-layout.md\":\"DuBu9a96\",\"reference_prompt.md\":\"DnaD5dNK\",\"reference_result.md\":\"BK5SxDrf\",\"reference_schedules.md\":\"DI_JrHgq\",\"reference_sessions.md\":\"CueyOHSL\",\"reference_skills.md\":\"l2ezAcaN\",\"reference_subagents.md\":\"CfsIloPm\",\"reference_tools.md\":\"aGrn99hP\",\"scaffolding-agents.md\":\"em43xlY1\",\"skills_ab.md\":\"CsFNatVx\",\"skills_create-agent.md\":\"BVoWPcan\",\"skills_debug.md\":\"CDbPhHfg\",\"skills_evals.md\":\"723kpUmA\",\"skills_framework-map.md\":\"haibFyoB\",\"skills_github.md\":\"D0JahM8c\",\"skills_hillclimb.md\":\"B_zJerxA\",\"skills_index.md\":\"DKwIxzGg\",\"skills_mcp-auth.md\":\"DGvFP3HE\",\"skills_otel.md\":\"CqhobH3B\",\"skills_setup-slack.md\":\"BBgx8lUz\",\"storage.md\":\"BOHeqk2M\",\"templates_agentic-owners.md\":\"BZSH4N9z\",\"templates_agents-md.md\":\"CXSMZljE\",\"templates_code-wiki.md\":\"CDYZIBFN\",\"templates_demo.md\":\"BeQX9V3H\",\"templates_grokbot-agents.md\":\"c1s6Flp9\",\"templates_pr-autofixer.md\":\"Bds_XWcV\",\"templates_security-help.md\":\"C3Ny_Qr2\",\"templates_security-reviewer.md\":\"lshxbCLK\",\"templates_triage.md\":\"Co4UNzkZ\",\"troubleshooting.md\":\"Cus_YZga\"}");window.__VP_SITE_DATA__=JSON.parse("{\"lang\":\"en-US\",\"dir\":\"ltr\",\"title\":\"Agent SDK\",\"description\":\"Filesystem-first framework for building and serving Cursor agents.\",\"base\":\"/docs/\",\"head\":[],\"router\":{\"prefetchLinks\":true},\"appearance\":true,\"themeConfig\":{\"sdkVersion\":\"0.1.105\",\"nav\":[{\"text\":\"Quickstart\",\"link\":\"/quickstart\"},{\"text\":\"Guides\",\"link\":\"/guides/webhooks\"},{\"text\":\"Templates\",\"link\":\"/templates/demo\"},{\"text\":\"Reference\",\"link\":\"/reference/project-layout\"}],\"search\":{\"provider\":\"local\"},\"outline\":{\"level\":[2,3]},\"sidebar\":[{\"text\":\"Getting started\",\"items\":[{\"text\":\"Overview\",\"link\":\"/\"},{\"text\":\"Quickstart\",\"link\":\"/quickstart\"},{\"text\":\"Scaffold an agent with Cursor\",\"link\":\"/scaffolding-agents\"},{\"text\":\"Convert a Cursor Automation\",\"link\":\"/guides/convert-automation\"},{\"text\":\"Concepts\",\"link\":\"/concepts\"}]},{\"text\":\"Templates\",\"items\":[{\"text\":\"Demo agent\",\"link\":\"/templates/demo\"},{\"text\":\"Grok Bot agents\",\"link\":\"/templates/grokbot-agents\"},{\"text\":\"Code wiki\",\"link\":\"/templates/code-wiki\"},{\"text\":\"Living AGENTS.md\",\"link\":\"/templates/agents-md\"},{\"text\":\"Security reviewer\",\"link\":\"/templates/security-reviewer\"},{\"text\":\"Security help\",\"link\":\"/templates/security-help\"},{\"text\":\"Agentic Owners\",\"link\":\"/templates/agentic-owners\"},{\"text\":\"PR autofixer\",\"link\":\"/templates/pr-autofixer\"},{\"text\":\"Triage agent\",\"link\":\"/templates/triage\"}]},{\"text\":\"Improving agents\",\"items\":[{\"text\":\"Building agents with agents\",\"link\":\"/building-with-agents\"},{\"text\":\"Evals\",\"link\":\"/evals\"},{\"text\":\"Live A/B metrics\",\"link\":\"/ab\"},{\"text\":\"Storage\",\"link\":\"/storage\"},{\"text\":\"Hillclimbing\",\"link\":\"/hillclimbing\"}]},{\"text\":\"Coding-agent skills\",\"collapsed\":true,\"items\":[{\"text\":\"Overview\",\"link\":\"/skills/\"},{\"text\":\"ab\",\"link\":\"/skills/ab\"},{\"text\":\"create-agent\",\"link\":\"/skills/create-agent\"},{\"text\":\"debug\",\"link\":\"/skills/debug\"},{\"text\":\"evals\",\"link\":\"/skills/evals\"},{\"text\":\"framework-map\",\"link\":\"/skills/framework-map\"},{\"text\":\"github\",\"link\":\"/skills/github\"},{\"text\":\"hillclimb\",\"link\":\"/skills/hillclimb\"},{\"text\":\"mcp-auth\",\"link\":\"/skills/mcp-auth\"},{\"text\":\"otel\",\"link\":\"/skills/otel\"},{\"text\":\"setup-slack\",\"link\":\"/skills/setup-slack\"}]},{\"text\":\"Guides\",\"items\":[{\"text\":\"Webhooks & custom channels\",\"link\":\"/guides/webhooks\"},{\"text\":\"GitHub\",\"link\":\"/guides/github\"},{\"text\":\"Slack\",\"link\":\"/guides/slack\"},{\"text\":\"Human-in-the-loop approvals\",\"link\":\"/guides/human-in-the-loop\"},{\"text\":\"Agent-to-agent\",\"link\":\"/guides/agent-to-agent\"},{\"text\":\"Host MCP OAuth\",\"link\":\"/guides/mcp-oauth\"},{\"text\":\"Cursor cloud agents\",\"link\":\"/guides/cloud-agents\"},{\"text\":\"Cursor Grok Bot agents\",\"link\":\"/guides/grokbot-agents\"},{\"text\":\"OpenTelemetry\",\"link\":\"/guides/opentelemetry\"}]},{\"text\":\"Operating\",\"items\":[{\"text\":\"Deployment\",\"link\":\"/deployment\"},{\"text\":\"Fix common problems\",\"link\":\"/troubleshooting\"}]},{\"text\":\"Reference\",\"collapsed\":false,\"items\":[{\"text\":\"Project layout\",\"link\":\"/reference/project-layout\"},{\"text\":\"Agent config\",\"link\":\"/reference/agent-config\"},{\"text\":\"Instructions\",\"link\":\"/reference/instructions\"},{\"text\":\"Tools\",\"link\":\"/reference/tools\"},{\"text\":\"prompt\",\"link\":\"/reference/prompt\"},{\"text\":\"Skills\",\"link\":\"/reference/skills\"},{\"text\":\"MCP Connections\",\"link\":\"/reference/connections\"},{\"text\":\"Subagents\",\"link\":\"/reference/subagents\"},{\"text\":\"Extensions\",\"link\":\"/reference/extensions\"},{\"text\":\"Channels\",\"link\":\"/reference/channels\"},{\"text\":\"Schedules & reminders\",\"link\":\"/reference/schedules\"},{\"text\":\"Hooks\",\"link\":\"/reference/hooks\"},{\"text\":\"Artifacts\",\"link\":\"/reference/artifacts\"},{\"text\":\"Turn result\",\"link\":\"/reference/result\"},{\"text\":\"Sessions & streaming\",\"link\":\"/reference/sessions\"},{\"text\":\"Playground\",\"link\":\"/reference/playground\"},{\"text\":\"CLI\",\"link\":\"/reference/cli\"},{\"text\":\"HTTP API\",\"link\":\"/reference/http-api\"}]}]},\"locales\":{},\"scrollOffset\":134,\"cleanUrls\":false}");</script>
107
+ <span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agent-sdk</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> status</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> evalrun_…</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --prod</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --slug</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> vulnerability-scanner</span></span></code></pre></div><h2 id="what-good-cases-assert" tabindex="-1">What good cases assert <a class="header-anchor" href="#what-good-cases-assert" aria-label="Permalink to &quot;What good cases assert&quot;">​</a></h2><p>Gate decisions and shape, not prose. Model wording varies run to run. Tool choice, tool avoidance, and output structure are the stable contract.</p><ol><li><code>t.succeeded()</code>: always, first.</li><li>The tool decision: <code>calledTool</code> for the intended path, <code>notCalledTool</code> for the likely wrong alternative. The pair is stronger than either alone.</li><li>Output shape: a regex for the contract (<code>/ready|blocked/i</code>, a JSON marker, a findings-block fence), never exact sentences.</li><li>For structured output, parse <code>t.reply</code> and check fields with <code>satisfies</code> instead of substring-matching JSON.</li></ol><p>The common failure modes: asserting exact phrasing, packing more than about five gates into one case (split it), and cases that depend on live external state that drifts (pin the input; see fixtures).</p><h2 id="pick-fixtures-by-agent-type" tabindex="-1">Pick fixtures by agent type <a class="header-anchor" href="#pick-fixtures-by-agent-type" aria-label="Permalink to &quot;Pick fixtures by agent type&quot;">​</a></h2><p>The right fixture depends on the surface under test.</p><table tabindex="0"><thead><tr><th>Agent surface</th><th>Fixture</th></tr></thead><tbody><tr><td>Chat / domain assistant</td><td>A canonical prompt string, chosen once and frozen</td></tr><tr><td>Tool-heavy</td><td>Run <code>agent-sdk call &lt;tool&gt;</code> first to pin what the tool returns, then freeze the prompt that triggers it</td></tr><tr><td>GitHub webhook</td><td><code>agent-sdk github replay &lt;pr&gt; --events &#39;*&#39; --dry-run --out fixtures/github</code> snapshots real payloads for offline replay (<a href="./guides/github.html">GitHub guide</a>)</td></tr><tr><td>PR reviewer with host preparation</td><td>Diff, metadata, and gold labels pinned to commit SHAs; keep any live PR matrix small</td></tr><tr><td>Workspace-dependent</td><td><code>workspaceFiles</code> in <code>t.send</code> options, never developer-machine paths</td></tr></tbody></table><p>Tag the fast, reliably passing core <code>smoke</code> and run <code>--tag smoke</code> in the inner loop. Leave slow or flaky-prone cases untagged for explicit runs.</p><h3 id="materialize-api-backed-fixtures" tabindex="-1">Materialize API-backed fixtures <a class="header-anchor" href="#materialize-api-backed-fixtures" aria-label="Permalink to &quot;Materialize API-backed fixtures&quot;">​</a></h3><p>An input that only points at external data, such as a pull request URL, snapshot id, or pair of commit SHAs, is not self-contained. Fetch it once and commit the rendered fixture before you expand the suite.</p><ol><li>Save the diff, metadata, and labels under <code>fixtures/</code> at pinned revisions.</li><li>Seed those files with <code>workspaceFiles</code>, or read them from the fixture directory.</li><li>Assert decisions and output shape against the saved evidence.</li><li>Keep a small <code>smoke</code> subset for any remaining live pipeline checks.</li></ol><p>Read committed fixtures with <code>@cursor/july/evals/loaders</code>: <code>loadJson</code>, <code>loadJsonl</code>, and <code>loadYaml</code> resolve relative paths against the project root the runner discovered, not the cwd the CLI was invoked from (<code>resolveFixturePath</code> and <code>evalFixtureRoot</code> expose the same resolution for other file formats).</p><p><code>maxConcurrency</code> limits parallel datapoints. It does not limit model or API fan-out inside one datapoint. Materialized fixtures prevent a large suite from exhausting provider and GitHub rate limits. The <a href="./skills/evals.html">evals skill</a> has the full fixture workflow.</p><h2 id="keep-improvements-with-regression-evals" tabindex="-1">Keep improvements with regression evals <a class="header-anchor" href="#keep-improvements-with-regression-evals" aria-label="Permalink to &quot;Keep improvements with regression evals&quot;">​</a></h2><p>Every <a href="./hillclimbing.html">hillclimb</a> round that keeps a change must land an eval that would have failed before the change. If you can&#39;t express the improvement as a gate (a <code>calledTool</code> shift, a bounded <code>action.result</code> count, an output-shape regex), the improvement is unverified, and it&#39;ll regress silently.</p><p>The rule cuts the other way too: never weaken an existing gate to make a round pass. That&#39;s the freeze line moving, and it turns your regression suite into a list of checks that no longer protect anything.</p><h2 id="compare-variants-on-live-traffic" tabindex="-1">Compare variants on live traffic <a class="header-anchor" href="#compare-variants-on-live-traffic" aria-label="Permalink to &quot;Compare variants on live traffic&quot;">​</a></h2><p>Use <code>defineAB</code> to compare variant metrics on live sessions. It is not a test runner and has no <code>agent-sdk ab</code> command. Keep <code>defineEval</code> as the regression ratchet. Eval sessions do not enroll or change live metrics. See <a href="./ab.html">Live A/B metrics</a> for assignment, behavior, collection, and inspection.</p><h2 id="what-s-next" tabindex="-1">What&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./ab.html">Live A/B metrics</a>: sticky variants and cumulative metrics on live sessions</li><li><a href="./hillclimbing.html">Hillclimbing</a>: the loop evals make trustworthy</li><li><a href="./building-with-agents.html">Building agents with agents</a>: have a coding agent write the first suite</li><li><a href="./guides/github.html">GitHub guide</a>: deterministic webhook fixtures with <code>github replay</code></li><li><a href="./reference/sessions.html">Sessions and streaming</a>: the events <code>t.events</code> contains</li></ul></div></div></main><footer class="VPDocFooter" data-v-38549ef7 data-v-215f93e6><!--[--><!--]--><!----><nav class="prev-next" aria-labelledby="doc-footer-aria-label" data-v-215f93e6><span class="visually-hidden" id="doc-footer-aria-label" data-v-215f93e6>Pager</span><div class="pager" data-v-215f93e6><a class="VPLink link pager-link prev" href="/docs/building-with-agents.html" data-v-215f93e6><!--[--><span class="desc" data-v-215f93e6>Previous page</span><span class="title" data-v-215f93e6>Building agents with agents</span><!--]--></a></div><div class="pager" data-v-215f93e6><a class="VPLink link pager-link next" href="/docs/ab.html" data-v-215f93e6><!--[--><span class="desc" data-v-215f93e6>Next page</span><span class="title" data-v-215f93e6>Live A/B metrics</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
108
+ <script>window.__VP_HASH_MAP__=JSON.parse("{\"ab.md\":\"mlVgqvSk\",\"building-with-agents.md\":\"CUSWxlP_\",\"concepts.md\":\"DgEcZOfT\",\"deployment.md\":\"MFPKqsqI\",\"evals.md\":\"CbMoebP1\",\"guides_agent-to-agent.md\":\"BDb0t1QV\",\"guides_cloud-agents.md\":\"DS8RIjwx\",\"guides_convert-automation.md\":\"B4sjlodG\",\"guides_github.md\":\"c0gdGJ-D\",\"guides_grokbot-agents.md\":\"DVqdhBKb\",\"guides_human-in-the-loop.md\":\"By1G2T3_\",\"guides_mcp-oauth.md\":\"Dp6cDP7f\",\"guides_opentelemetry.md\":\"SCRMskno\",\"guides_slack.md\":\"9oHPye9o\",\"guides_webhooks.md\":\"DKdA43Qm\",\"hillclimbing.md\":\"CpTGTCle\",\"index.md\":\"BW_6tOgR\",\"quickstart.md\":\"DdQOF7Y8\",\"reference_agent-config.md\":\"CHNpiyp4\",\"reference_artifacts.md\":\"Dior32Qw\",\"reference_channels.md\":\"D-qTqwcq\",\"reference_cli.md\":\"Dm67hd2D\",\"reference_connections.md\":\"Di6jJAXF\",\"reference_extensions.md\":\"CGmMLblt\",\"reference_hooks.md\":\"Ddt5DdgJ\",\"reference_http-api.md\":\"oySXBO8o\",\"reference_instructions.md\":\"CR7XSsGk\",\"reference_playground.md\":\"4myJPxrf\",\"reference_project-layout.md\":\"DuBu9a96\",\"reference_prompt.md\":\"DnaD5dNK\",\"reference_result.md\":\"BK5SxDrf\",\"reference_schedules.md\":\"DI_JrHgq\",\"reference_sessions.md\":\"CueyOHSL\",\"reference_skills.md\":\"l2ezAcaN\",\"reference_subagents.md\":\"CfsIloPm\",\"reference_tools.md\":\"aGrn99hP\",\"scaffolding-agents.md\":\"em43xlY1\",\"skills_ab.md\":\"CsFNatVx\",\"skills_create-agent.md\":\"BVoWPcan\",\"skills_debug.md\":\"CDbPhHfg\",\"skills_evals.md\":\"723kpUmA\",\"skills_framework-map.md\":\"haibFyoB\",\"skills_github.md\":\"D0JahM8c\",\"skills_hillclimb.md\":\"B_zJerxA\",\"skills_index.md\":\"DKwIxzGg\",\"skills_mcp-auth.md\":\"DGvFP3HE\",\"skills_otel.md\":\"B5Ddv1xH\",\"skills_setup-slack.md\":\"BBgx8lUz\",\"storage.md\":\"BOHeqk2M\",\"templates_agentic-owners.md\":\"BZSH4N9z\",\"templates_agents-md.md\":\"CXSMZljE\",\"templates_code-wiki.md\":\"CDYZIBFN\",\"templates_demo.md\":\"BeQX9V3H\",\"templates_grokbot-agents.md\":\"c1s6Flp9\",\"templates_pr-autofixer.md\":\"Bds_XWcV\",\"templates_security-help.md\":\"C3Ny_Qr2\",\"templates_security-reviewer.md\":\"lshxbCLK\",\"templates_triage.md\":\"Co4UNzkZ\",\"troubleshooting.md\":\"Cus_YZga\"}");window.__VP_SITE_DATA__=JSON.parse("{\"lang\":\"en-US\",\"dir\":\"ltr\",\"title\":\"Agent SDK\",\"description\":\"Filesystem-first framework for building and serving Cursor agents.\",\"base\":\"/docs/\",\"head\":[],\"router\":{\"prefetchLinks\":true},\"appearance\":true,\"themeConfig\":{\"sdkVersion\":\"0.1.107\",\"nav\":[{\"text\":\"Quickstart\",\"link\":\"/quickstart\"},{\"text\":\"Guides\",\"link\":\"/guides/webhooks\"},{\"text\":\"Templates\",\"link\":\"/templates/demo\"},{\"text\":\"Reference\",\"link\":\"/reference/project-layout\"}],\"search\":{\"provider\":\"local\"},\"outline\":{\"level\":[2,3]},\"sidebar\":[{\"text\":\"Getting started\",\"items\":[{\"text\":\"Overview\",\"link\":\"/\"},{\"text\":\"Quickstart\",\"link\":\"/quickstart\"},{\"text\":\"Scaffold an agent with Cursor\",\"link\":\"/scaffolding-agents\"},{\"text\":\"Convert a Cursor Automation\",\"link\":\"/guides/convert-automation\"},{\"text\":\"Concepts\",\"link\":\"/concepts\"}]},{\"text\":\"Templates\",\"items\":[{\"text\":\"Demo agent\",\"link\":\"/templates/demo\"},{\"text\":\"Grok Bot agents\",\"link\":\"/templates/grokbot-agents\"},{\"text\":\"Code wiki\",\"link\":\"/templates/code-wiki\"},{\"text\":\"Living AGENTS.md\",\"link\":\"/templates/agents-md\"},{\"text\":\"Security reviewer\",\"link\":\"/templates/security-reviewer\"},{\"text\":\"Security help\",\"link\":\"/templates/security-help\"},{\"text\":\"Agentic Owners\",\"link\":\"/templates/agentic-owners\"},{\"text\":\"PR autofixer\",\"link\":\"/templates/pr-autofixer\"},{\"text\":\"Triage agent\",\"link\":\"/templates/triage\"}]},{\"text\":\"Improving agents\",\"items\":[{\"text\":\"Building agents with agents\",\"link\":\"/building-with-agents\"},{\"text\":\"Evals\",\"link\":\"/evals\"},{\"text\":\"Live A/B metrics\",\"link\":\"/ab\"},{\"text\":\"Storage\",\"link\":\"/storage\"},{\"text\":\"Hillclimbing\",\"link\":\"/hillclimbing\"}]},{\"text\":\"Coding-agent skills\",\"collapsed\":true,\"items\":[{\"text\":\"Overview\",\"link\":\"/skills/\"},{\"text\":\"ab\",\"link\":\"/skills/ab\"},{\"text\":\"create-agent\",\"link\":\"/skills/create-agent\"},{\"text\":\"debug\",\"link\":\"/skills/debug\"},{\"text\":\"evals\",\"link\":\"/skills/evals\"},{\"text\":\"framework-map\",\"link\":\"/skills/framework-map\"},{\"text\":\"github\",\"link\":\"/skills/github\"},{\"text\":\"hillclimb\",\"link\":\"/skills/hillclimb\"},{\"text\":\"mcp-auth\",\"link\":\"/skills/mcp-auth\"},{\"text\":\"otel\",\"link\":\"/skills/otel\"},{\"text\":\"setup-slack\",\"link\":\"/skills/setup-slack\"}]},{\"text\":\"Guides\",\"items\":[{\"text\":\"Webhooks & custom channels\",\"link\":\"/guides/webhooks\"},{\"text\":\"GitHub\",\"link\":\"/guides/github\"},{\"text\":\"Slack\",\"link\":\"/guides/slack\"},{\"text\":\"Human-in-the-loop approvals\",\"link\":\"/guides/human-in-the-loop\"},{\"text\":\"Agent-to-agent\",\"link\":\"/guides/agent-to-agent\"},{\"text\":\"Host MCP OAuth\",\"link\":\"/guides/mcp-oauth\"},{\"text\":\"Cursor cloud agents\",\"link\":\"/guides/cloud-agents\"},{\"text\":\"Cursor Grok Bot agents\",\"link\":\"/guides/grokbot-agents\"},{\"text\":\"OpenTelemetry\",\"link\":\"/guides/opentelemetry\"}]},{\"text\":\"Operating\",\"items\":[{\"text\":\"Deployment\",\"link\":\"/deployment\"},{\"text\":\"Fix common problems\",\"link\":\"/troubleshooting\"}]},{\"text\":\"Reference\",\"collapsed\":false,\"items\":[{\"text\":\"Project layout\",\"link\":\"/reference/project-layout\"},{\"text\":\"Agent config\",\"link\":\"/reference/agent-config\"},{\"text\":\"Instructions\",\"link\":\"/reference/instructions\"},{\"text\":\"Tools\",\"link\":\"/reference/tools\"},{\"text\":\"prompt\",\"link\":\"/reference/prompt\"},{\"text\":\"Skills\",\"link\":\"/reference/skills\"},{\"text\":\"MCP Connections\",\"link\":\"/reference/connections\"},{\"text\":\"Subagents\",\"link\":\"/reference/subagents\"},{\"text\":\"Extensions\",\"link\":\"/reference/extensions\"},{\"text\":\"Channels\",\"link\":\"/reference/channels\"},{\"text\":\"Schedules & reminders\",\"link\":\"/reference/schedules\"},{\"text\":\"Hooks\",\"link\":\"/reference/hooks\"},{\"text\":\"Artifacts\",\"link\":\"/reference/artifacts\"},{\"text\":\"Turn result\",\"link\":\"/reference/result\"},{\"text\":\"Sessions & streaming\",\"link\":\"/reference/sessions\"},{\"text\":\"Playground\",\"link\":\"/reference/playground\"},{\"text\":\"CLI\",\"link\":\"/reference/cli\"},{\"text\":\"HTTP API\",\"link\":\"/reference/http-api\"}]}]},\"locales\":{},\"scrollOffset\":134,\"cleanUrls\":false}");</script>
109
109
 
110
110
  </body>
111
111
  </html>