@cursor/july 0.1.10 → 0.1.11
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.
- package/AGENTS.md +1 -0
- package/dist/bin/agent-serve.js +17 -3
- package/dist/channels/github/github-channel.d.ts.map +1 -1
- package/dist/channels/github/github-channel.js +2 -1
- package/dist/channels/github/types.d.ts +4 -0
- package/dist/channels/github/types.d.ts.map +1 -1
- package/dist/channels/slack/cursor-account.d.ts.map +1 -1
- package/dist/channels/slack/cursor-account.js +1 -1
- package/dist/channels/slack/dispatch.d.ts +13 -0
- package/dist/channels/slack/dispatch.d.ts.map +1 -1
- package/dist/channels/slack/dispatch.js +36 -11
- package/dist/channels/slack/eval-directive.d.ts +63 -0
- package/dist/channels/slack/eval-directive.d.ts.map +1 -0
- package/dist/channels/slack/eval-directive.js +283 -0
- package/dist/channels/slack/index.d.ts +1 -0
- package/dist/channels/slack/index.d.ts.map +1 -1
- package/dist/channels/slack/index.js +1 -0
- package/dist/channels/slack/post-update-delivery.js +318 -0
- package/dist/channels/slack/slack-channel.d.ts.map +1 -1
- package/dist/channels/slack/slack-channel.js +4 -2
- package/dist/channels/slack/types.d.ts +7 -0
- package/dist/channels/slack/types.d.ts.map +1 -1
- package/dist/connections.d.ts +10 -5
- package/dist/connections.d.ts.map +1 -1
- package/dist/connections.js +10 -5
- package/dist/docs/404.html +2 -2
- package/dist/docs/ab.html +4 -4
- package/dist/docs/assets/{app.CY2buaEE.js → app.BcXUGca7.js} +1 -1
- package/dist/docs/assets/chunks/@localSearchIndexroot.BIbvK8M1.js +1 -0
- package/dist/docs/assets/chunks/{VPLocalSearchBox.Ix954SFJ.js → VPLocalSearchBox.CWur6XGZ.js} +1 -1
- package/dist/docs/assets/chunks/{theme.PDwuJbyi.js → theme.BnvulOX2.js} +2 -2
- package/dist/docs/assets/{deployment.md.DC5-ImcM.js → deployment.md.TecHo0_2.js} +3 -3
- package/dist/docs/assets/{deployment.md.DC5-ImcM.lean.js → deployment.md.TecHo0_2.lean.js} +1 -1
- package/dist/docs/assets/guides_mcp-oauth.md.DqmEOQY1.js +29 -0
- package/dist/docs/assets/guides_mcp-oauth.md.DqmEOQY1.lean.js +1 -0
- package/dist/docs/assets/{index.md.COiu-1jL.js → index.md.t0TM2Qzz.js} +3 -3
- package/dist/docs/assets/{index.md.COiu-1jL.lean.js → index.md.t0TM2Qzz.lean.js} +1 -1
- package/dist/docs/assets/{reference_cli.md.CjfiV2Q-.js → reference_cli.md.B74NrVzb.js} +3 -3
- package/dist/docs/assets/{reference_cli.md.CjfiV2Q-.lean.js → reference_cli.md.B74NrVzb.lean.js} +1 -1
- package/dist/docs/assets/{reference_connections.md.FbWupyzW.js → reference_connections.md.CsSr3rDz.js} +8 -3
- package/dist/docs/assets/{reference_connections.md.FbWupyzW.lean.js → reference_connections.md.CsSr3rDz.lean.js} +1 -1
- package/dist/docs/assets/reference_prompt.md.BaiweQxE.js +1 -0
- package/dist/docs/assets/reference_prompt.md.BaiweQxE.lean.js +1 -0
- package/dist/docs/assets/{reference_tools.md.CqgJroI0.js → reference_tools.md.BRSDnTbN.js} +16 -3
- package/dist/docs/assets/{reference_tools.md.CqgJroI0.lean.js → reference_tools.md.BRSDnTbN.lean.js} +1 -1
- package/dist/docs/assets/{troubleshooting.md.DkovjS7h.js → troubleshooting.md.CmQkmnzC.js} +1 -1
- package/dist/docs/assets/{troubleshooting.md.DkovjS7h.lean.js → troubleshooting.md.CmQkmnzC.lean.js} +1 -1
- package/dist/docs/building-with-agents.html +4 -4
- package/dist/docs/concepts.html +4 -4
- package/dist/docs/deployment.html +7 -7
- package/dist/docs/evals.html +4 -4
- package/dist/docs/example-agents/approval-buddy.html +4 -4
- package/dist/docs/example-agents/benny.html +4 -4
- package/dist/docs/example-agents/bugbot.html +4 -4
- package/dist/docs/example-agents/codebase-wiki.html +4 -4
- package/dist/docs/example-agents/codeowners-review.html +4 -4
- package/dist/docs/example-agents/concierge.html +4 -4
- package/dist/docs/example-agents/fsd.html +4 -4
- package/dist/docs/example-agents/index.html +4 -4
- package/dist/docs/example-agents/knowledge-base.html +4 -4
- package/dist/docs/example-agents/oncall.html +4 -4
- package/dist/docs/example-agents/security-reviewer.html +4 -4
- package/dist/docs/example-agents/slack-agent.html +4 -4
- package/dist/docs/example-agents/weather-agent.html +4 -4
- package/dist/docs/guides/agent-to-agent.html +4 -4
- package/dist/docs/guides/cloud-runtime.html +4 -4
- package/dist/docs/guides/github.html +4 -4
- package/dist/docs/guides/human-in-the-loop.html +4 -4
- package/dist/docs/guides/mcp-oauth.html +53 -0
- package/dist/docs/guides/slack.html +4 -4
- package/dist/docs/guides/webhooks.html +4 -4
- package/dist/docs/hashmap.json +1 -1
- package/dist/docs/hillclimbing.html +4 -4
- package/dist/docs/index.html +6 -6
- package/dist/docs/quickstart.html +4 -4
- package/dist/docs/reference/agent-config.html +4 -4
- package/dist/docs/reference/channels.html +4 -4
- package/dist/docs/reference/cli.html +6 -6
- package/dist/docs/reference/connections.html +12 -7
- package/dist/docs/reference/hooks.html +4 -4
- package/dist/docs/reference/http-api.html +4 -4
- package/dist/docs/reference/instructions.html +4 -4
- package/dist/docs/reference/playground.html +4 -4
- package/dist/docs/reference/project-layout.html +4 -4
- package/dist/docs/reference/prompt.html +33 -0
- package/dist/docs/reference/schedules.html +4 -4
- package/dist/docs/reference/sessions.html +4 -4
- package/dist/docs/reference/skills.html +5 -5
- package/dist/docs/reference/subagents.html +4 -4
- package/dist/docs/reference/tools.html +20 -7
- package/dist/docs/scaffolding-agents.html +4 -4
- package/dist/docs/storage.html +4 -4
- package/dist/docs/troubleshooting.html +5 -5
- package/dist/evals.d.ts +39 -0
- package/dist/evals.d.ts.map +1 -1
- package/dist/index.d.ts +3 -1
- package/dist/index.d.ts.map +1 -1
- package/dist/index.js +2 -1
- package/dist/internal/cli-mcp-oauth.d.ts +16 -0
- package/dist/internal/cli-mcp-oauth.d.ts.map +1 -0
- package/dist/internal/cli-mcp-oauth.js +194 -0
- package/dist/internal/cli-mcp.d.ts +4 -1
- package/dist/internal/cli-mcp.d.ts.map +1 -1
- package/dist/internal/cli-mcp.js +19 -1
- package/dist/internal/corp-ca.d.ts +9 -0
- package/dist/internal/corp-ca.d.ts.map +1 -0
- package/dist/internal/corp-ca.js +44 -0
- package/dist/internal/cursor/account-mcp.d.ts +17 -4
- package/dist/internal/cursor/account-mcp.d.ts.map +1 -1
- package/dist/internal/cursor/account-mcp.js +105 -17
- package/dist/internal/discovery.d.ts.map +1 -1
- package/dist/internal/discovery.js +64 -10
- package/dist/internal/distribution.d.ts.map +1 -1
- package/dist/internal/distribution.js +1 -0
- package/dist/internal/json-dir-store.js +100 -0
- package/dist/internal/logs-client.js +1 -1
- package/dist/internal/mcp-host.d.ts.map +1 -1
- package/dist/internal/mcp-host.js +49 -26
- package/dist/internal/mcp-oauth.d.ts +88 -0
- package/dist/internal/mcp-oauth.d.ts.map +1 -0
- package/dist/internal/mcp-oauth.js +385 -0
- package/dist/internal/resolved-connections.d.ts.map +1 -1
- package/dist/internal/resolved-connections.js +4 -1
- package/dist/internal/sdk-runner.d.ts +9 -0
- package/dist/internal/sdk-runner.d.ts.map +1 -1
- package/dist/internal/sdk-runner.js +28 -2
- package/dist/internal/server.d.ts.map +1 -1
- package/dist/internal/server.js +13 -5
- package/dist/internal/session-engine.d.ts +7 -9
- package/dist/internal/session-engine.d.ts.map +1 -1
- package/dist/internal/session-engine.js +96 -64
- package/dist/internal/sessions-client.d.ts.map +1 -1
- package/dist/internal/sessions-client.js +2 -1
- package/dist/internal/stream-progress.d.ts.map +1 -1
- package/dist/internal/stream-progress.js +8 -4
- package/dist/internal/turn-governor.d.ts +54 -0
- package/dist/internal/turn-governor.d.ts.map +1 -0
- package/dist/internal/turn-governor.js +111 -0
- package/dist/playground/assets/{index-C0-QEG4O.js → index-BIGLubwK.js} +46 -46
- package/dist/playground/assets/{index-z3JY9O2x.css → index-C17ZQKDG.css} +1 -1
- package/dist/playground/index.html +2 -2
- package/dist/prompt.d.ts +39 -0
- package/dist/prompt.d.ts.map +1 -0
- package/dist/prompt.js +70 -0
- package/dist/types.d.ts +78 -5
- package/dist/types.d.ts.map +1 -1
- package/dist/types.js +6 -0
- package/docs/.vitepress/cache/deps/package.json +3 -0
- package/docs/README.md +4 -0
- package/docs/deployment.md +6 -1
- package/docs/guides/mcp-oauth.md +157 -0
- package/docs/reference/cli.md +28 -0
- package/docs/reference/connections.md +28 -0
- package/docs/reference/prompt.md +47 -0
- package/docs/reference/tools.md +21 -0
- package/docs/troubleshooting.md +11 -0
- package/package.json +9 -2
- package/skills/create-agent/SKILL.md +7 -3
- package/skills/framework-map/SKILL.md +1 -0
- package/skills/mcp-auth/SKILL.md +138 -0
- package/src/bin/agent-serve.ts +20 -3
- package/src/channels/github/github-channel.ts +2 -0
- package/src/channels/github/types.ts +4 -0
- package/src/channels/slack/cursor-account.ts +2 -0
- package/src/channels/slack/dispatch.ts +49 -0
- package/src/channels/slack/eval-directive.ts +348 -0
- package/src/channels/slack/index.ts +1 -0
- package/src/channels/slack/slack-channel.ts +3 -0
- package/src/channels/slack/types.ts +7 -0
- package/src/connections.ts +10 -5
- package/src/evals.ts +43 -0
- package/src/index.ts +3 -0
- package/src/internal/cli-mcp-oauth.ts +254 -0
- package/src/internal/cli-mcp.ts +25 -2
- package/src/internal/cursor/account-mcp.ts +153 -41
- package/src/internal/discovery.ts +70 -8
- package/src/internal/distribution.ts +1 -0
- package/src/internal/logs-client.ts +1 -1
- package/src/internal/mcp-host.ts +33 -9
- package/src/internal/mcp-oauth.ts +496 -0
- package/src/internal/resolved-connections.ts +6 -1
- package/src/internal/sdk-runner.ts +37 -1
- package/src/internal/server.ts +13 -5
- package/src/internal/session-engine.ts +115 -74
- package/src/internal/sessions-client.ts +2 -1
- package/src/internal/stream-progress.ts +8 -0
- package/src/internal/turn-governor.ts +145 -0
- package/src/prompt.ts +83 -0
- package/src/types.ts +76 -6
- package/dist/docs/assets/chunks/@localSearchIndexroot.OfDZ5CXg.js +0 -1
- package/dist/internal/persistence-coordinator.d.ts +0 -127
- package/dist/internal/persistence-coordinator.d.ts.map +0 -1
- package/dist/persistence.d.ts +0 -184
- package/dist/persistence.d.ts.map +0 -1
package/dist/docs/evals.html
CHANGED
|
@@ -9,16 +9,16 @@
|
|
|
9
9
|
<link rel="preload stylesheet" href="/docs/assets/style.Bj-IXa5W.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.
|
|
12
|
+
<script type="module" src="/docs/assets/app.BcXUGca7.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.
|
|
14
|
+
<link rel="modulepreload" href="/docs/assets/chunks/theme.BnvulOX2.js">
|
|
15
15
|
<link rel="modulepreload" href="/docs/assets/chunks/framework.CAZyNGu9.js">
|
|
16
16
|
<link rel="modulepreload" href="/docs/assets/evals.md.DAgEc_hL.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>
|
|
18
18
|
<script id="check-mac-os">document.documentElement.classList.toggle("mac",/Mac|iPhone|iPod|iPad/i.test(navigator.platform));</script>
|
|
19
19
|
</head>
|
|
20
20
|
<body>
|
|
21
|
-
<div id="app"><div class="Layout" data-v-05085060><!--[--><!--]--><!--[--><span tabindex="-1" data-v-30dc752a></span><a href="#VPContent" class="VPSkipLink visually-hidden" data-v-30dc752a>Skip to content</a><!--]--><!----><header class="VPNav" data-v-05085060 data-v-cbe59881><div class="VPNavBar" data-v-cbe59881 data-v-dd8a1864><div class="wrapper" data-v-dd8a1864><div class="container" data-v-dd8a1864><div class="title" data-v-dd8a1864><div class="VPNavBarTitle has-sidebar" data-v-dd8a1864 data-v-f9346b2b><a class="title" href="/docs/" data-v-f9346b2b><!--[--><!--]--><!----><span data-v-f9346b2b>agentkit</span><!--[--><!--]--></a></div></div><div class="content" data-v-dd8a1864><div class="content-body" data-v-dd8a1864><!--[--><!--]--><div class="VPNavBarSearch search" data-v-dd8a1864><!--[--><!----><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-dd8a1864 data-v-7ed6e5ac><span id="main-nav-aria-label" class="visually-hidden" data-v-7ed6e5ac> Main Navigation </span><!--[--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/quickstart.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Quickstart</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/guides/webhooks.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Guides</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/example-agents/" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Examples</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/reference/project-layout.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Reference</span><!--]--></a><!--]--><!--]--></nav><!----><div class="VPNavBarAppearance appearance" data-v-dd8a1864 data-v-2b2b395e><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-2b2b395e data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></span><!--]--></span></span></button></div><!----><div class="VPFlyout VPNavBarExtra extra" data-v-dd8a1864 data-v-f3b584b7 data-v-82133de6><button type="button" class="button" aria-haspopup="true" aria-expanded="false" aria-label="extra navigation" data-v-82133de6><span class="vpi-more-horizontal icon" data-v-82133de6></span></button><div class="menu" data-v-82133de6><div class="VPMenu" data-v-82133de6 data-v-759c9e19><!----><!--[--><!--[--><!----><div class="group" data-v-f3b584b7><div class="item appearance" data-v-f3b584b7><p class="label" data-v-f3b584b7>Appearance</p><div class="appearance-action" data-v-f3b584b7><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-f3b584b7 data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></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-dd8a1864 data-v-ab07e844><span class="container" data-v-ab07e844><span class="top" data-v-ab07e844></span><span class="middle" data-v-ab07e844></span><span class="bottom" data-v-ab07e844></span></span></button></div></div></div></div><div class="divider" data-v-dd8a1864><div class="divider-line" data-v-dd8a1864></div></div></div><!----></header><div class="VPLocalNav has-sidebar empty" data-v-05085060 data-v-826b8761><div class="container" data-v-826b8761><button class="menu" aria-expanded="false" aria-controls="VPSidebarNav" data-v-826b8761><span class="vpi-align-left menu-icon" data-v-826b8761></span><span class="menu-text" data-v-826b8761>Menu</span></button><div class="VPLocalNavOutlineDropdown" style="--vp-vh:0px;" data-v-826b8761 data-v-01af9e73><button data-v-01af9e73>Return to top</button><!----></div></div></div><aside class="VPSidebar" data-v-05085060 data-v-5cd5b4d8><div class="curtain" data-v-5cd5b4d8></div><nav class="nav" id="VPSidebarNav" aria-labelledby="sidebar-aria-label" tabindex="-1" data-v-5cd5b4d8><span class="visually-hidden" id="sidebar-aria-label" data-v-5cd5b4d8> Sidebar Navigation </span><!--[--><!--]--><!--[--><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Getting started</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/quickstart.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Quickstart</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/scaffolding-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Scaffold an agent with Cursor</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/concepts.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concepts</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 has-active" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Improving agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/building-with-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Building agents with agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/evals.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/ab.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Live A/B metrics</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/storage.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Storage</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/hillclimbing.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hillclimbing</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Guides</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/webhooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Webhooks & custom channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/github.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>GitHub</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/slack.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/human-in-the-loop.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Human-in-the-loop approvals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/agent-to-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent-to-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/cloud-runtime.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Cloud runtime</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Example agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Choose an example</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/weather-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Weather agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/slack-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/concierge.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concierge</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/benny.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playbook router</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/oncall.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Alert investigator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/bugbot.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>PR evidence reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/approval-buddy.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Approval Buddy</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/security-reviewer.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Security Reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/fsd.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Remote PR coordinator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/knowledge-base.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Knowledge base</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codebase-wiki.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codebase wiki</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codeowners-review.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codeowners review</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Operating</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/deployment.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Deployment</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/troubleshooting.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Fix common problems</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 collapsible" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Reference</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-b6893b1b><span class="vpi-chevron-right caret-icon" data-v-b6893b1b></span></div></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/project-layout.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Project layout</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/agent-config.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent config</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/instructions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Instructions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/tools.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Tools</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/skills.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Skills</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/connections.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>MCP Connections</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/subagents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Subagents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/channels.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/schedules.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Schedules & reminders</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/hooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hooks</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/sessions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Sessions & streaming</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/playground.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playground</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/cli.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>CLI</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/http-api.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>HTTP API</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><!--]--><!--[--><!--]--></nav></aside><div class="VPContent has-sidebar" id="VPContent" data-v-05085060 data-v-8eedf2ac><div class="VPDoc has-sidebar has-aside" data-v-8eedf2ac data-v-53b9fda6><!--[--><!--]--><div class="container" data-v-53b9fda6><div class="aside" data-v-53b9fda6><div class="aside-curtain" data-v-53b9fda6></div><div class="aside-container" data-v-53b9fda6><div class="aside-content" data-v-53b9fda6><div class="VPDocAside" data-v-53b9fda6 data-v-e448ecab><!--[--><!--]--><!--[--><!--]--><nav aria-labelledby="doc-outline-aria-label" class="VPDocAsideOutline" data-v-e448ecab data-v-b0b42374><div class="content" data-v-b0b42374><div class="outline-marker" data-v-b0b42374></div><div aria-level="2" class="outline-title" id="doc-outline-aria-label" role="heading" data-v-b0b42374>On this page</div><ul class="VPDocOutlineItem root" data-v-b0b42374 data-v-ae03bd52><!--[--><!--]--></ul></div></nav><!--[--><!--]--><div class="spacer" data-v-e448ecab></div><!--[--><!--]--><!----><!--[--><!--]--><!--[--><!--]--></div></div></div></div><div class="content" data-v-53b9fda6><div class="content-container" data-v-53b9fda6><!--[--><!--]--><main class="main" data-v-53b9fda6><div style="position:relative;" class="vp-doc _docs_evals" data-v-53b9fda6><div><h1 id="evals" tabindex="-1">Evals <a class="header-anchor" href="#evals" aria-label="Permalink to "Evals""></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't regress the agent, and last month'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><div class="note custom-block github-alert"><p class="custom-block-title">NOTE</p><p>Import paths here use <code>@cursor/july/evals</code>. On projects still using <code>@cursor/july</code>, swap the import and run <code>agent-serve eval</code>. See <a href="./README.html#run-the-cli">Run the CLI</a> for the full rename table.</p></div><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 "Define evals with `defineEval`""></a></h2><p>agentkit discovers evals under the project-root <code>evals/</code> directory, in <code>.eval.ts</code> or <code>.eval.js</code> files. That'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's identity, so you don'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>
|
|
21
|
+
<div id="app"><div class="Layout" data-v-05085060><!--[--><!--]--><!--[--><span tabindex="-1" data-v-30dc752a></span><a href="#VPContent" class="VPSkipLink visually-hidden" data-v-30dc752a>Skip to content</a><!--]--><!----><header class="VPNav" data-v-05085060 data-v-cbe59881><div class="VPNavBar" data-v-cbe59881 data-v-dd8a1864><div class="wrapper" data-v-dd8a1864><div class="container" data-v-dd8a1864><div class="title" data-v-dd8a1864><div class="VPNavBarTitle has-sidebar" data-v-dd8a1864 data-v-f9346b2b><a class="title" href="/docs/" data-v-f9346b2b><!--[--><!--]--><!----><span data-v-f9346b2b>agentkit</span><!--[--><!--]--></a></div></div><div class="content" data-v-dd8a1864><div class="content-body" data-v-dd8a1864><!--[--><!--]--><div class="VPNavBarSearch search" data-v-dd8a1864><!--[--><!----><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-dd8a1864 data-v-7ed6e5ac><span id="main-nav-aria-label" class="visually-hidden" data-v-7ed6e5ac> Main Navigation </span><!--[--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/quickstart.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Quickstart</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/guides/webhooks.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Guides</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/example-agents/" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Examples</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/reference/project-layout.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Reference</span><!--]--></a><!--]--><!--]--></nav><!----><div class="VPNavBarAppearance appearance" data-v-dd8a1864 data-v-2b2b395e><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-2b2b395e data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></span><!--]--></span></span></button></div><!----><div class="VPFlyout VPNavBarExtra extra" data-v-dd8a1864 data-v-f3b584b7 data-v-82133de6><button type="button" class="button" aria-haspopup="true" aria-expanded="false" aria-label="extra navigation" data-v-82133de6><span class="vpi-more-horizontal icon" data-v-82133de6></span></button><div class="menu" data-v-82133de6><div class="VPMenu" data-v-82133de6 data-v-759c9e19><!----><!--[--><!--[--><!----><div class="group" data-v-f3b584b7><div class="item appearance" data-v-f3b584b7><p class="label" data-v-f3b584b7>Appearance</p><div class="appearance-action" data-v-f3b584b7><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-f3b584b7 data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></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-dd8a1864 data-v-ab07e844><span class="container" data-v-ab07e844><span class="top" data-v-ab07e844></span><span class="middle" data-v-ab07e844></span><span class="bottom" data-v-ab07e844></span></span></button></div></div></div></div><div class="divider" data-v-dd8a1864><div class="divider-line" data-v-dd8a1864></div></div></div><!----></header><div class="VPLocalNav has-sidebar empty" data-v-05085060 data-v-826b8761><div class="container" data-v-826b8761><button class="menu" aria-expanded="false" aria-controls="VPSidebarNav" data-v-826b8761><span class="vpi-align-left menu-icon" data-v-826b8761></span><span class="menu-text" data-v-826b8761>Menu</span></button><div class="VPLocalNavOutlineDropdown" style="--vp-vh:0px;" data-v-826b8761 data-v-01af9e73><button data-v-01af9e73>Return to top</button><!----></div></div></div><aside class="VPSidebar" data-v-05085060 data-v-5cd5b4d8><div class="curtain" data-v-5cd5b4d8></div><nav class="nav" id="VPSidebarNav" aria-labelledby="sidebar-aria-label" tabindex="-1" data-v-5cd5b4d8><span class="visually-hidden" id="sidebar-aria-label" data-v-5cd5b4d8> Sidebar Navigation </span><!--[--><!--]--><!--[--><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Getting started</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/quickstart.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Quickstart</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/scaffolding-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Scaffold an agent with Cursor</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/concepts.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concepts</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 has-active" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Improving agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/building-with-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Building agents with agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/evals.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/ab.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Live A/B metrics</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/storage.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Storage</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/hillclimbing.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hillclimbing</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Guides</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/webhooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Webhooks & custom channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/github.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>GitHub</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/slack.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/human-in-the-loop.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Human-in-the-loop approvals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/agent-to-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent-to-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/cloud-runtime.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Cloud runtime</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Example agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Choose an example</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/weather-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Weather agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/slack-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/concierge.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concierge</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/benny.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playbook router</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/oncall.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Alert investigator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/bugbot.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>PR evidence reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/approval-buddy.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Approval Buddy</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/security-reviewer.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Security Reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/fsd.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Remote PR coordinator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/knowledge-base.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Knowledge base</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codebase-wiki.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codebase wiki</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codeowners-review.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codeowners review</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Operating</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/deployment.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Deployment</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/troubleshooting.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Fix common problems</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 collapsible" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Reference</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-b6893b1b><span class="vpi-chevron-right caret-icon" data-v-b6893b1b></span></div></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/project-layout.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Project layout</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/agent-config.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent config</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/instructions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Instructions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/tools.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Tools</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/prompt.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>prompt</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/skills.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Skills</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/connections.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>MCP Connections</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/subagents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Subagents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/channels.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/schedules.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Schedules & reminders</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/hooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hooks</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/sessions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Sessions & streaming</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/playground.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playground</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/cli.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>CLI</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/http-api.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>HTTP API</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><!--]--><!--[--><!--]--></nav></aside><div class="VPContent has-sidebar" id="VPContent" data-v-05085060 data-v-8eedf2ac><div class="VPDoc has-sidebar has-aside" data-v-8eedf2ac data-v-53b9fda6><!--[--><!--]--><div class="container" data-v-53b9fda6><div class="aside" data-v-53b9fda6><div class="aside-curtain" data-v-53b9fda6></div><div class="aside-container" data-v-53b9fda6><div class="aside-content" data-v-53b9fda6><div class="VPDocAside" data-v-53b9fda6 data-v-e448ecab><!--[--><!--]--><!--[--><!--]--><nav aria-labelledby="doc-outline-aria-label" class="VPDocAsideOutline" data-v-e448ecab data-v-b0b42374><div class="content" data-v-b0b42374><div class="outline-marker" data-v-b0b42374></div><div aria-level="2" class="outline-title" id="doc-outline-aria-label" role="heading" data-v-b0b42374>On this page</div><ul class="VPDocOutlineItem root" data-v-b0b42374 data-v-ae03bd52><!--[--><!--]--></ul></div></nav><!--[--><!--]--><div class="spacer" data-v-e448ecab></div><!--[--><!--]--><!----><!--[--><!--]--><!--[--><!--]--></div></div></div></div><div class="content" data-v-53b9fda6><div class="content-container" data-v-53b9fda6><!--[--><!--]--><main class="main" data-v-53b9fda6><div style="position:relative;" class="vp-doc _docs_evals" data-v-53b9fda6><div><h1 id="evals" tabindex="-1">Evals <a class="header-anchor" href="#evals" aria-label="Permalink to "Evals""></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't regress the agent, and last month'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><div class="note custom-block github-alert"><p class="custom-block-title">NOTE</p><p>Import paths here use <code>@cursor/july/evals</code>. On projects still using <code>@cursor/july</code>, swap the import and run <code>agent-serve eval</code>. See <a href="./README.html#run-the-cli">Run the CLI</a> for the full rename table.</p></div><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 "Define evals with `defineEval`""></a></h2><p>agentkit discovers evals under the project-root <code>evals/</code> directory, in <code>.eval.ts</code> or <code>.eval.js</code> files. That'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's identity, so you don'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>
|
|
22
22
|
<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;"> "@cursor/july/evals"</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">;</span></span>
|
|
23
23
|
<span class="line"></span>
|
|
24
24
|
<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>
|
|
@@ -103,7 +103,7 @@
|
|
|
103
103
|
<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> }</span></span>
|
|
104
104
|
<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> ]</span></span>
|
|
105
105
|
<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">}</span></span></code></pre></div><p>Each case result can also include <code>description</code>, <code>finalText</code>, <code>tools</code>, <code>error</code>, and tool arguments or output. This shape lets CI report the failed assertion without parsing terminal text.</p><h2 id="run-evals-in-the-playground" tabindex="-1">Run evals in the playground <a class="header-anchor" href="#run-evals-in-the-playground" aria-label="Permalink to "Run evals in the playground""></a></h2><p>Start the server with <code>--dev</code>, open the playground, and choose <strong>Evals</strong>. You can run every case or one case, watch progress, and open the resulting session trace.</p><div class="language-bash vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">bash</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> serve</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dir</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> .</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dev</span></span></code></pre></div><p>Playground runs target the live server instead of an ephemeral one. Their sessions appear in the session list. One eval batch can run at a time. By default those batches are <strong>in-memory only</strong> (capped by <code>maxPlaygroundRuns</code>); set <code>persistRuns</code> in <code>evals.config.ts</code> if you need them after a serve restart — see <a href="#configure-eval-runs">Configure eval runs</a>.</p><p>The UI uses the playground eval routes (available without <code>--dev</code>): <code>GET /v1/dev/evals</code> lists datapoints and config (includes <code>maxPlaygroundRuns</code> / <code>durableRuns</code>), <code>GET /v1/dev/evals/runs</code> rehydrates recent batches after navigation, <code>POST /v1/dev/evals/runs</code> starts a batch, and <code>GET /v1/dev/evals/runs/:runId</code> polls it. See <a href="./reference/http-api.html#playground-eval-routes">Playground eval routes</a>. The start request returns <code>202</code> while cases run in the background. Poll until the snapshot status becomes <code>completed</code> or <code>failed</code>. Configuration errors appear on a failed snapshot.</p><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 "What good cases assert""></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 "Pick fixtures by agent type""></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>agentkit call <tool></code> first to pin what the tool returns, then freeze the prompt that triggers it</td></tr><tr><td>GitHub webhook</td><td><code>agentkit github replay <pr> --events '*' --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 "Materialize API-backed fixtures""></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><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/SKILL.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 "Keep improvements with regression evals""></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'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'll regress silently.</p><p>The rule cuts the other way too: never weaken an existing gate to make a round pass. That'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 "Compare variants on live traffic""></a></h2><p>Use <code>defineAB</code> to compare variant metrics on live sessions. It is not a test runner and has no <code>agentkit 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's next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to "What's next""></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-53b9fda6 data-v-3958cc3f><!--[--><!--]--><!----><nav class="prev-next" aria-labelledby="doc-footer-aria-label" data-v-3958cc3f><span class="visually-hidden" id="doc-footer-aria-label" data-v-3958cc3f>Pager</span><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link prev" href="/docs/building-with-agents.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Previous page</span><span class="title" data-v-3958cc3f>Building agents with agents</span><!--]--></a></div><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link next" href="/docs/ab.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Next page</span><span class="title" data-v-3958cc3f>Live A/B metrics</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
|
|
106
|
-
<script>window.__VP_HASH_MAP__=JSON.parse("{\"ab.md\":\"BMCZ6Hd7\",\"building-with-agents.md\":\"CJCtZCyi\",\"concepts.md\":\"Cfb9b-k1\",\"deployment.md\":\"
|
|
106
|
+
<script>window.__VP_HASH_MAP__=JSON.parse("{\"ab.md\":\"BMCZ6Hd7\",\"building-with-agents.md\":\"CJCtZCyi\",\"concepts.md\":\"Cfb9b-k1\",\"deployment.md\":\"TecHo0_2\",\"evals.md\":\"DAgEc_hL\",\"example-agents_approval-buddy.md\":\"DFGBYLcc\",\"example-agents_benny.md\":\"B0gjhI-p\",\"example-agents_bugbot.md\":\"DelIdhxB\",\"example-agents_codebase-wiki.md\":\"DC6sgwn0\",\"example-agents_codeowners-review.md\":\"Ku_tG2RY\",\"example-agents_concierge.md\":\"4rQTSMXt\",\"example-agents_fsd.md\":\"CzgUrDfi\",\"example-agents_index.md\":\"BPXjUtRB\",\"example-agents_knowledge-base.md\":\"BPJiVueF\",\"example-agents_oncall.md\":\"BG_sUMly\",\"example-agents_security-reviewer.md\":\"CE9v-aKm\",\"example-agents_slack-agent.md\":\"buLbgvBf\",\"example-agents_weather-agent.md\":\"C9Qv-W0o\",\"guides_agent-to-agent.md\":\"Bpzgq2Pq\",\"guides_cloud-runtime.md\":\"gVzabdQL\",\"guides_github.md\":\"DOOCpqsW\",\"guides_human-in-the-loop.md\":\"DlUqsp1S\",\"guides_mcp-oauth.md\":\"DqmEOQY1\",\"guides_slack.md\":\"FWale6Ip\",\"guides_webhooks.md\":\"B1EswtUu\",\"hillclimbing.md\":\"DHNast08\",\"index.md\":\"t0TM2Qzz\",\"quickstart.md\":\"BU6Iwi_9\",\"reference_agent-config.md\":\"DrW2JUM8\",\"reference_channels.md\":\"DdmiKgqf\",\"reference_cli.md\":\"B74NrVzb\",\"reference_connections.md\":\"CsSr3rDz\",\"reference_hooks.md\":\"DyLVfE1O\",\"reference_http-api.md\":\"Dx_nmDG6\",\"reference_instructions.md\":\"CgoV-YEb\",\"reference_playground.md\":\"CPZhfYaO\",\"reference_project-layout.md\":\"Bd_CKtNS\",\"reference_prompt.md\":\"BaiweQxE\",\"reference_schedules.md\":\"w_F2mXB6\",\"reference_sessions.md\":\"DLd6mvbv\",\"reference_skills.md\":\"B_jHN7JL\",\"reference_subagents.md\":\"zWAMNfi1\",\"reference_tools.md\":\"BRSDnTbN\",\"scaffolding-agents.md\":\"C3pTrmoE\",\"storage.md\":\"DRTdnFvd\",\"troubleshooting.md\":\"CmQkmnzC\"}");window.__VP_SITE_DATA__=JSON.parse("{\"lang\":\"en-US\",\"dir\":\"ltr\",\"title\":\"agentkit\",\"description\":\"Filesystem-first framework for building and serving Cursor agents.\",\"base\":\"/docs/\",\"head\":[],\"router\":{\"prefetchLinks\":true},\"appearance\":true,\"themeConfig\":{\"nav\":[{\"text\":\"Quickstart\",\"link\":\"/quickstart\"},{\"text\":\"Guides\",\"link\":\"/guides/webhooks\"},{\"text\":\"Examples\",\"link\":\"/example-agents/\"},{\"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\":\"Concepts\",\"link\":\"/concepts\"}]},{\"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\":\"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\":\"Cloud runtime\",\"link\":\"/guides/cloud-runtime\"}]},{\"text\":\"Example agents\",\"items\":[{\"text\":\"Choose an example\",\"link\":\"/example-agents/\"},{\"text\":\"Weather agent\",\"link\":\"/example-agents/weather-agent\"},{\"text\":\"Slack agent\",\"link\":\"/example-agents/slack-agent\"},{\"text\":\"Concierge\",\"link\":\"/example-agents/concierge\"},{\"text\":\"Playbook router\",\"link\":\"/example-agents/benny\"},{\"text\":\"Alert investigator\",\"link\":\"/example-agents/oncall\"},{\"text\":\"PR evidence reviewer\",\"link\":\"/example-agents/bugbot\"},{\"text\":\"Approval Buddy\",\"link\":\"/example-agents/approval-buddy\"},{\"text\":\"Security Reviewer\",\"link\":\"/example-agents/security-reviewer\"},{\"text\":\"Remote PR coordinator\",\"link\":\"/example-agents/fsd\"},{\"text\":\"Knowledge base\",\"link\":\"/example-agents/knowledge-base\"},{\"text\":\"Codebase wiki\",\"link\":\"/example-agents/codebase-wiki\"},{\"text\":\"Codeowners review\",\"link\":\"/example-agents/codeowners-review\"}]},{\"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\":\"Channels\",\"link\":\"/reference/channels\"},{\"text\":\"Schedules & reminders\",\"link\":\"/reference/schedules\"},{\"text\":\"Hooks\",\"link\":\"/reference/hooks\"},{\"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
107
|
|
|
108
108
|
</body>
|
|
109
109
|
</html>
|
|
@@ -9,16 +9,16 @@
|
|
|
9
9
|
<link rel="preload stylesheet" href="/docs/assets/style.Bj-IXa5W.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.
|
|
12
|
+
<script type="module" src="/docs/assets/app.BcXUGca7.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.
|
|
14
|
+
<link rel="modulepreload" href="/docs/assets/chunks/theme.BnvulOX2.js">
|
|
15
15
|
<link rel="modulepreload" href="/docs/assets/chunks/framework.CAZyNGu9.js">
|
|
16
16
|
<link rel="modulepreload" href="/docs/assets/example-agents_approval-buddy.md.DFGBYLcc.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>
|
|
18
18
|
<script id="check-mac-os">document.documentElement.classList.toggle("mac",/Mac|iPhone|iPod|iPad/i.test(navigator.platform));</script>
|
|
19
19
|
</head>
|
|
20
20
|
<body>
|
|
21
|
-
<div id="app"><div class="Layout" data-v-05085060><!--[--><!--]--><!--[--><span tabindex="-1" data-v-30dc752a></span><a href="#VPContent" class="VPSkipLink visually-hidden" data-v-30dc752a>Skip to content</a><!--]--><!----><header class="VPNav" data-v-05085060 data-v-cbe59881><div class="VPNavBar" data-v-cbe59881 data-v-dd8a1864><div class="wrapper" data-v-dd8a1864><div class="container" data-v-dd8a1864><div class="title" data-v-dd8a1864><div class="VPNavBarTitle has-sidebar" data-v-dd8a1864 data-v-f9346b2b><a class="title" href="/docs/" data-v-f9346b2b><!--[--><!--]--><!----><span data-v-f9346b2b>agentkit</span><!--[--><!--]--></a></div></div><div class="content" data-v-dd8a1864><div class="content-body" data-v-dd8a1864><!--[--><!--]--><div class="VPNavBarSearch search" data-v-dd8a1864><!--[--><!----><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-dd8a1864 data-v-7ed6e5ac><span id="main-nav-aria-label" class="visually-hidden" data-v-7ed6e5ac> Main Navigation </span><!--[--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/quickstart.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Quickstart</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/guides/webhooks.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Guides</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/example-agents/" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Examples</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/reference/project-layout.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Reference</span><!--]--></a><!--]--><!--]--></nav><!----><div class="VPNavBarAppearance appearance" data-v-dd8a1864 data-v-2b2b395e><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-2b2b395e data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></span><!--]--></span></span></button></div><!----><div class="VPFlyout VPNavBarExtra extra" data-v-dd8a1864 data-v-f3b584b7 data-v-82133de6><button type="button" class="button" aria-haspopup="true" aria-expanded="false" aria-label="extra navigation" data-v-82133de6><span class="vpi-more-horizontal icon" data-v-82133de6></span></button><div class="menu" data-v-82133de6><div class="VPMenu" data-v-82133de6 data-v-759c9e19><!----><!--[--><!--[--><!----><div class="group" data-v-f3b584b7><div class="item appearance" data-v-f3b584b7><p class="label" data-v-f3b584b7>Appearance</p><div class="appearance-action" data-v-f3b584b7><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-f3b584b7 data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></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-dd8a1864 data-v-ab07e844><span class="container" data-v-ab07e844><span class="top" data-v-ab07e844></span><span class="middle" data-v-ab07e844></span><span class="bottom" data-v-ab07e844></span></span></button></div></div></div></div><div class="divider" data-v-dd8a1864><div class="divider-line" data-v-dd8a1864></div></div></div><!----></header><div class="VPLocalNav has-sidebar empty" data-v-05085060 data-v-826b8761><div class="container" data-v-826b8761><button class="menu" aria-expanded="false" aria-controls="VPSidebarNav" data-v-826b8761><span class="vpi-align-left menu-icon" data-v-826b8761></span><span class="menu-text" data-v-826b8761>Menu</span></button><div class="VPLocalNavOutlineDropdown" style="--vp-vh:0px;" data-v-826b8761 data-v-01af9e73><button data-v-01af9e73>Return to top</button><!----></div></div></div><aside class="VPSidebar" data-v-05085060 data-v-5cd5b4d8><div class="curtain" data-v-5cd5b4d8></div><nav class="nav" id="VPSidebarNav" aria-labelledby="sidebar-aria-label" tabindex="-1" data-v-5cd5b4d8><span class="visually-hidden" id="sidebar-aria-label" data-v-5cd5b4d8> Sidebar Navigation </span><!--[--><!--]--><!--[--><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Getting started</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/quickstart.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Quickstart</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/scaffolding-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Scaffold an agent with Cursor</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/concepts.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concepts</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Improving agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/building-with-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Building agents with agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/evals.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/ab.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Live A/B metrics</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/storage.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Storage</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/hillclimbing.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hillclimbing</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Guides</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/webhooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Webhooks & custom channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/github.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>GitHub</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/slack.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/human-in-the-loop.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Human-in-the-loop approvals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/agent-to-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent-to-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/cloud-runtime.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Cloud runtime</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 has-active" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Example agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Choose an example</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/weather-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Weather agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/slack-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/concierge.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concierge</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/benny.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playbook router</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/oncall.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Alert investigator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/bugbot.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>PR evidence reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/approval-buddy.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Approval Buddy</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/security-reviewer.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Security Reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/fsd.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Remote PR coordinator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/knowledge-base.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Knowledge base</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codebase-wiki.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codebase wiki</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codeowners-review.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codeowners review</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Operating</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/deployment.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Deployment</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/troubleshooting.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Fix common problems</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 collapsible" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Reference</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-b6893b1b><span class="vpi-chevron-right caret-icon" data-v-b6893b1b></span></div></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/project-layout.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Project layout</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/agent-config.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent config</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/instructions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Instructions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/tools.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Tools</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/skills.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Skills</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/connections.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>MCP Connections</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/subagents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Subagents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/channels.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/schedules.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Schedules & reminders</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/hooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hooks</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/sessions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Sessions & streaming</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/playground.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playground</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/cli.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>CLI</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/http-api.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>HTTP API</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><!--]--><!--[--><!--]--></nav></aside><div class="VPContent has-sidebar" id="VPContent" data-v-05085060 data-v-8eedf2ac><div class="VPDoc has-sidebar has-aside" data-v-8eedf2ac data-v-53b9fda6><!--[--><!--]--><div class="container" data-v-53b9fda6><div class="aside" data-v-53b9fda6><div class="aside-curtain" data-v-53b9fda6></div><div class="aside-container" data-v-53b9fda6><div class="aside-content" data-v-53b9fda6><div class="VPDocAside" data-v-53b9fda6 data-v-e448ecab><!--[--><!--]--><!--[--><!--]--><nav aria-labelledby="doc-outline-aria-label" class="VPDocAsideOutline" data-v-e448ecab data-v-b0b42374><div class="content" data-v-b0b42374><div class="outline-marker" data-v-b0b42374></div><div aria-level="2" class="outline-title" id="doc-outline-aria-label" role="heading" data-v-b0b42374>On this page</div><ul class="VPDocOutlineItem root" data-v-b0b42374 data-v-ae03bd52><!--[--><!--]--></ul></div></nav><!--[--><!--]--><div class="spacer" data-v-e448ecab></div><!--[--><!--]--><!----><!--[--><!--]--><!--[--><!--]--></div></div></div></div><div class="content" data-v-53b9fda6><div class="content-container" data-v-53b9fda6><!--[--><!--]--><main class="main" data-v-53b9fda6><div style="position:relative;" class="vp-doc _docs_example-agents_approval-buddy" data-v-53b9fda6><div><h1 id="keep-pr-approval-policy-deterministic-with-approval-buddy" tabindex="-1">Keep PR approval policy deterministic with Approval Buddy <a class="header-anchor" href="#keep-pr-approval-policy-deterministic-with-approval-buddy" aria-label="Permalink to "Keep PR approval policy deterministic with Approval Buddy""></a></h1><p>Approval Buddy approves eligible pull requests from a fixed roster and declines every other request. GitHub still blocks self-approval when the stamp identity authored the PR. Code decides eligibility. The model prepares evidence, runs two specialist reviews, and passes their findings to the approval tool without changing the policy decision.</p><p>Use this example when an agent can make a judgment inside a workflow, but authorization and the final side effect must stay in deterministic code.</p><p><a href="./../../examples/approval-buddy/">Browse the Approval Buddy source.</a></p><h2 id="keep-approval-policy-in-code" tabindex="-1">Keep approval policy in code <a class="header-anchor" href="#keep-approval-policy-in-code" aria-label="Permalink to "Keep approval policy in code""></a></h2><p>Approval Buddy draws three hard boundaries:</p><ul><li><code>prepare_review</code> and <code>approve_pr</code> re-read the live PR and apply the same eligibility rules.</li><li>Two subagents inspect prepared evidence, but their findings never grant or block approval.</li><li>Only <code>approve_pr</code> posts the GitHub review.</li></ul><p>A spoofed webhook, Slack message, or model claim can't add someone to the buddy roster. The mutating tool checks the source of truth immediately before it acts.</p><h2 id="follow-the-intended-stamp-flow" tabindex="-1">Follow the intended stamp flow <a class="header-anchor" href="#follow-the-intended-stamp-flow" aria-label="Permalink to "Follow the intended stamp flow""></a></h2><p>The root instructions ask the model to run this sequence for a qualifying PR:</p><ol><li>A non-draft <code>pull_request</code> event arrives with action <code>opened</code>, <code>reopened</code>, or <code>ready_for_review</code>.</li><li>The GitHub channel checks its repository allowlist and starts a session.</li><li><code>turn.started</code> posts a pending commit status.</li><li>The model calls <code>prepare_review</code>.</li><li>Host code fetches the live PR. It checks the author, open state, merged state, and draft state.</li><li>A qualifying PR gets <code>pr/MANIFEST.md</code>, <code>pr/meta.json</code>, and <code>pr/diff.patch</code> in the session workspace. Diffs above 2,000,000 characters are truncated and marked in metadata.</li><li>The model calls both review subagents through the built-in <code>task</code> tool.</li><li>It concatenates their contracted replies and calls <code>approve_pr</code>.</li><li><code>approve_pr</code> re-runs eligibility, posts an <code>APPROVE</code> review, and returns the outcome.</li><li>The channel posts a final commit status. A self-approval block also gets a short timeline comment because no approval review can appear.</li></ol><p>Ineligible PRs skip evidence and subagents. The model still calls <code>approve_pr</code> so the deterministic tool returns the formal decline reason.</p><p>Steps 4 through 9 are prompt-driven. The channel doesn't enforce tool order or prove both subagents ran, and <code>approve_pr</code> accepts missing findings. A failed turn clears the pending status with a green non-blocking result without approving the PR.</p><h2 id="map-the-framework-features" tabindex="-1">Map the framework features <a class="header-anchor" href="#map-the-framework-features" aria-label="Permalink to "Map the framework features""></a></h2><table tabindex="0"><thead><tr><th>Capability</th><th>Source</th><th>Role</th></tr></thead><tbody><tr><td>Root agent and policy prompt</td><td><a href="../../examples/approval-buddy/agent/agent.ts"><code>agent/agent.ts</code></a>, <a href="./../../examples/approval-buddy/agent/instructions.html"><code>agent/instructions.md</code></a></td><td>Configure the local agent and describe orchestration order.</td></tr><tr><td>GitHub channel</td><td><a href="../../examples/approval-buddy/agent/channels/github.ts"><code>agent/channels/github.ts</code></a></td><td>Filter wakes, lease GitHub access, and publish status events.</td></tr><tr><td>Slack channel</td><td><a href="../../examples/approval-buddy/agent/channels/slack.ts"><code>agent/channels/slack.ts</code></a></td><td>Accept approval-bot stamp and qualification requests.</td></tr><tr><td>Server tools</td><td><a href="./../../examples/approval-buddy/agent/tools/"><code>agent/tools/</code></a></td><td>Prepare evidence, approve, list buddies, and search GIFs.</td></tr><tr><td>Deterministic policy</td><td><a href="../../examples/approval-buddy/agent/lib/approve.ts"><code>agent/lib/approve.ts</code></a>, <a href="../../examples/approval-buddy/agent/lib/buddies.ts"><code>agent/lib/buddies.ts</code></a></td><td>Own the roster and live eligibility checks.</td></tr><tr><td>Review subagents</td><td><a href="./../../examples/approval-buddy/agent/subagents/"><code>agent/subagents/</code></a></td><td>Run deep audit and code-quality passes over the same evidence.</td></tr><tr><td>Storage</td><td><a href="../../examples/approval-buddy/agent/storage.ts"><code>agent/storage.ts</code></a></td><td>Persist sessions and events with <code>cursorHostedStorage</code> (Bugbot <code>agent_serve_*</code>).</td></tr><tr><td>Evals and unit tests</td><td><a href="./../../examples/approval-buddy/evals/"><code>evals/</code></a>, <a href="./../../examples/approval-buddy/agent/lib/"><code>agent/lib/</code></a></td><td>Protect routing, output contracts, policy, and GitHub behavior.</td></tr></tbody></table><p>There are no authored skills, MCP connections, schedules, reminders, hooks, A/B experiments, sandbox seeds, or tool approvals.</p><h2 id="prepare-credentials" tabindex="-1">Prepare credentials <a class="header-anchor" href="#prepare-credentials" aria-label="Permalink to "Prepare credentials""></a></h2><p>You need:</p><ul><li>Node 22.13 or newer.</li><li>An agent-runtime credential.</li><li>GitHub access to read PRs, post reviews, create commit statuses, and post the self-approval visibility comment.</li></ul><p>Optional GIF selection uses:</p><ul><li><code>GIPHY_API_KEY</code> or <code>APPROVAL_BUDDY_GIPHY_API_KEY</code>,</li><li><code>APPROVAL_BUDDY_STAMP_GIF</code>, or</li><li>severity-specific <code>APPROVAL_BUDDY_STAMP_GIF_<LEVEL></code> variables.</li></ul><p>If you enable Giphy in a hosted copy, declare its secret and <code>api.giphy.com</code> egress.</p><h2 id="validate-without-approving-a-pr" tabindex="-1">Validate without approving a PR <a class="header-anchor" href="#validate-without-approving-a-pr" aria-label="Permalink to "Validate without approving a PR""></a></h2><div class="language-bash vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">bash</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> validate</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dir</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> examples/approval-buddy</span></span>
|
|
21
|
+
<div id="app"><div class="Layout" data-v-05085060><!--[--><!--]--><!--[--><span tabindex="-1" data-v-30dc752a></span><a href="#VPContent" class="VPSkipLink visually-hidden" data-v-30dc752a>Skip to content</a><!--]--><!----><header class="VPNav" data-v-05085060 data-v-cbe59881><div class="VPNavBar" data-v-cbe59881 data-v-dd8a1864><div class="wrapper" data-v-dd8a1864><div class="container" data-v-dd8a1864><div class="title" data-v-dd8a1864><div class="VPNavBarTitle has-sidebar" data-v-dd8a1864 data-v-f9346b2b><a class="title" href="/docs/" data-v-f9346b2b><!--[--><!--]--><!----><span data-v-f9346b2b>agentkit</span><!--[--><!--]--></a></div></div><div class="content" data-v-dd8a1864><div class="content-body" data-v-dd8a1864><!--[--><!--]--><div class="VPNavBarSearch search" data-v-dd8a1864><!--[--><!----><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-dd8a1864 data-v-7ed6e5ac><span id="main-nav-aria-label" class="visually-hidden" data-v-7ed6e5ac> Main Navigation </span><!--[--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/quickstart.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Quickstart</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/guides/webhooks.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Guides</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/example-agents/" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Examples</span><!--]--></a><!--]--><!--[--><a class="VPLink link VPNavBarMenuLink" href="/docs/reference/project-layout.html" tabindex="0" data-v-7ed6e5ac data-v-65720067><!--[--><span data-v-65720067>Reference</span><!--]--></a><!--]--><!--]--></nav><!----><div class="VPNavBarAppearance appearance" data-v-dd8a1864 data-v-2b2b395e><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-2b2b395e data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></span><!--]--></span></span></button></div><!----><div class="VPFlyout VPNavBarExtra extra" data-v-dd8a1864 data-v-f3b584b7 data-v-82133de6><button type="button" class="button" aria-haspopup="true" aria-expanded="false" aria-label="extra navigation" data-v-82133de6><span class="vpi-more-horizontal icon" data-v-82133de6></span></button><div class="menu" data-v-82133de6><div class="VPMenu" data-v-82133de6 data-v-759c9e19><!----><!--[--><!--[--><!----><div class="group" data-v-f3b584b7><div class="item appearance" data-v-f3b584b7><p class="label" data-v-f3b584b7>Appearance</p><div class="appearance-action" data-v-f3b584b7><button class="VPSwitch VPSwitchAppearance" type="button" role="switch" title aria-checked="false" data-v-f3b584b7 data-v-182deabf data-v-2562b538><span class="check" data-v-2562b538><span class="icon" data-v-2562b538><!--[--><span class="vpi-sun sun" data-v-182deabf></span><span class="vpi-moon moon" data-v-182deabf></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-dd8a1864 data-v-ab07e844><span class="container" data-v-ab07e844><span class="top" data-v-ab07e844></span><span class="middle" data-v-ab07e844></span><span class="bottom" data-v-ab07e844></span></span></button></div></div></div></div><div class="divider" data-v-dd8a1864><div class="divider-line" data-v-dd8a1864></div></div></div><!----></header><div class="VPLocalNav has-sidebar empty" data-v-05085060 data-v-826b8761><div class="container" data-v-826b8761><button class="menu" aria-expanded="false" aria-controls="VPSidebarNav" data-v-826b8761><span class="vpi-align-left menu-icon" data-v-826b8761></span><span class="menu-text" data-v-826b8761>Menu</span></button><div class="VPLocalNavOutlineDropdown" style="--vp-vh:0px;" data-v-826b8761 data-v-01af9e73><button data-v-01af9e73>Return to top</button><!----></div></div></div><aside class="VPSidebar" data-v-05085060 data-v-5cd5b4d8><div class="curtain" data-v-5cd5b4d8></div><nav class="nav" id="VPSidebarNav" aria-labelledby="sidebar-aria-label" tabindex="-1" data-v-5cd5b4d8><span class="visually-hidden" id="sidebar-aria-label" data-v-5cd5b4d8> Sidebar Navigation </span><!--[--><!--]--><!--[--><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Getting started</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Overview</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/quickstart.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Quickstart</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/scaffolding-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Scaffold an agent with Cursor</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/concepts.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concepts</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Improving agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/building-with-agents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Building agents with agents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/evals.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Evals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/ab.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Live A/B metrics</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/storage.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Storage</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/hillclimbing.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hillclimbing</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Guides</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/webhooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Webhooks & custom channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/github.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>GitHub</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/slack.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/human-in-the-loop.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Human-in-the-loop approvals</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/agent-to-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent-to-agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/guides/cloud-runtime.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Cloud runtime</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 has-active" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Example agents</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Choose an example</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/weather-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Weather agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/slack-agent.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Slack agent</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/concierge.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Concierge</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/benny.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playbook router</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/oncall.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Alert investigator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/bugbot.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>PR evidence reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/approval-buddy.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Approval Buddy</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/security-reviewer.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Security Reviewer</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/fsd.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Remote PR coordinator</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/knowledge-base.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Knowledge base</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codebase-wiki.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codebase wiki</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/example-agents/codeowners-review.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Codeowners review</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Operating</h2><!----></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/deployment.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Deployment</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/troubleshooting.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Fix common problems</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><div class="no-transition group" data-v-6b69f472><section class="VPSidebarItem level-0 collapsible" data-v-6b69f472 data-v-b6893b1b><div class="item" role="button" tabindex="0" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><h2 class="text" data-v-b6893b1b>Reference</h2><div class="caret" role="button" aria-label="toggle section" tabindex="0" data-v-b6893b1b><span class="vpi-chevron-right caret-icon" data-v-b6893b1b></span></div></div><div class="items" data-v-b6893b1b><!--[--><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/project-layout.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Project layout</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/agent-config.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Agent config</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/instructions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Instructions</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/tools.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Tools</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/prompt.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>prompt</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/skills.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Skills</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/connections.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>MCP Connections</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/subagents.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Subagents</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/channels.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Channels</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/schedules.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Schedules & reminders</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/hooks.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Hooks</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/sessions.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Sessions & streaming</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/playground.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>Playground</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/cli.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>CLI</p><!--]--></a><!----></div><!----></div><div class="VPSidebarItem level-1 is-link" data-v-b6893b1b><div class="item" data-v-b6893b1b><div class="indicator" data-v-b6893b1b></div><a class="VPLink link link" href="/docs/reference/http-api.html" data-v-b6893b1b><!--[--><p class="text" data-v-b6893b1b>HTTP API</p><!--]--></a><!----></div><!----></div><!--]--></div></section></div><!--]--><!--[--><!--]--></nav></aside><div class="VPContent has-sidebar" id="VPContent" data-v-05085060 data-v-8eedf2ac><div class="VPDoc has-sidebar has-aside" data-v-8eedf2ac data-v-53b9fda6><!--[--><!--]--><div class="container" data-v-53b9fda6><div class="aside" data-v-53b9fda6><div class="aside-curtain" data-v-53b9fda6></div><div class="aside-container" data-v-53b9fda6><div class="aside-content" data-v-53b9fda6><div class="VPDocAside" data-v-53b9fda6 data-v-e448ecab><!--[--><!--]--><!--[--><!--]--><nav aria-labelledby="doc-outline-aria-label" class="VPDocAsideOutline" data-v-e448ecab data-v-b0b42374><div class="content" data-v-b0b42374><div class="outline-marker" data-v-b0b42374></div><div aria-level="2" class="outline-title" id="doc-outline-aria-label" role="heading" data-v-b0b42374>On this page</div><ul class="VPDocOutlineItem root" data-v-b0b42374 data-v-ae03bd52><!--[--><!--]--></ul></div></nav><!--[--><!--]--><div class="spacer" data-v-e448ecab></div><!--[--><!--]--><!----><!--[--><!--]--><!--[--><!--]--></div></div></div></div><div class="content" data-v-53b9fda6><div class="content-container" data-v-53b9fda6><!--[--><!--]--><main class="main" data-v-53b9fda6><div style="position:relative;" class="vp-doc _docs_example-agents_approval-buddy" data-v-53b9fda6><div><h1 id="keep-pr-approval-policy-deterministic-with-approval-buddy" tabindex="-1">Keep PR approval policy deterministic with Approval Buddy <a class="header-anchor" href="#keep-pr-approval-policy-deterministic-with-approval-buddy" aria-label="Permalink to "Keep PR approval policy deterministic with Approval Buddy""></a></h1><p>Approval Buddy approves eligible pull requests from a fixed roster and declines every other request. GitHub still blocks self-approval when the stamp identity authored the PR. Code decides eligibility. The model prepares evidence, runs two specialist reviews, and passes their findings to the approval tool without changing the policy decision.</p><p>Use this example when an agent can make a judgment inside a workflow, but authorization and the final side effect must stay in deterministic code.</p><p><a href="./../../examples/approval-buddy/">Browse the Approval Buddy source.</a></p><h2 id="keep-approval-policy-in-code" tabindex="-1">Keep approval policy in code <a class="header-anchor" href="#keep-approval-policy-in-code" aria-label="Permalink to "Keep approval policy in code""></a></h2><p>Approval Buddy draws three hard boundaries:</p><ul><li><code>prepare_review</code> and <code>approve_pr</code> re-read the live PR and apply the same eligibility rules.</li><li>Two subagents inspect prepared evidence, but their findings never grant or block approval.</li><li>Only <code>approve_pr</code> posts the GitHub review.</li></ul><p>A spoofed webhook, Slack message, or model claim can't add someone to the buddy roster. The mutating tool checks the source of truth immediately before it acts.</p><h2 id="follow-the-intended-stamp-flow" tabindex="-1">Follow the intended stamp flow <a class="header-anchor" href="#follow-the-intended-stamp-flow" aria-label="Permalink to "Follow the intended stamp flow""></a></h2><p>The root instructions ask the model to run this sequence for a qualifying PR:</p><ol><li>A non-draft <code>pull_request</code> event arrives with action <code>opened</code>, <code>reopened</code>, or <code>ready_for_review</code>.</li><li>The GitHub channel checks its repository allowlist and starts a session.</li><li><code>turn.started</code> posts a pending commit status.</li><li>The model calls <code>prepare_review</code>.</li><li>Host code fetches the live PR. It checks the author, open state, merged state, and draft state.</li><li>A qualifying PR gets <code>pr/MANIFEST.md</code>, <code>pr/meta.json</code>, and <code>pr/diff.patch</code> in the session workspace. Diffs above 2,000,000 characters are truncated and marked in metadata.</li><li>The model calls both review subagents through the built-in <code>task</code> tool.</li><li>It concatenates their contracted replies and calls <code>approve_pr</code>.</li><li><code>approve_pr</code> re-runs eligibility, posts an <code>APPROVE</code> review, and returns the outcome.</li><li>The channel posts a final commit status. A self-approval block also gets a short timeline comment because no approval review can appear.</li></ol><p>Ineligible PRs skip evidence and subagents. The model still calls <code>approve_pr</code> so the deterministic tool returns the formal decline reason.</p><p>Steps 4 through 9 are prompt-driven. The channel doesn't enforce tool order or prove both subagents ran, and <code>approve_pr</code> accepts missing findings. A failed turn clears the pending status with a green non-blocking result without approving the PR.</p><h2 id="map-the-framework-features" tabindex="-1">Map the framework features <a class="header-anchor" href="#map-the-framework-features" aria-label="Permalink to "Map the framework features""></a></h2><table tabindex="0"><thead><tr><th>Capability</th><th>Source</th><th>Role</th></tr></thead><tbody><tr><td>Root agent and policy prompt</td><td><a href="../../examples/approval-buddy/agent/agent.ts"><code>agent/agent.ts</code></a>, <a href="./../../examples/approval-buddy/agent/instructions.html"><code>agent/instructions.md</code></a></td><td>Configure the local agent and describe orchestration order.</td></tr><tr><td>GitHub channel</td><td><a href="../../examples/approval-buddy/agent/channels/github.ts"><code>agent/channels/github.ts</code></a></td><td>Filter wakes, lease GitHub access, and publish status events.</td></tr><tr><td>Slack channel</td><td><a href="../../examples/approval-buddy/agent/channels/slack.ts"><code>agent/channels/slack.ts</code></a></td><td>Accept approval-bot stamp and qualification requests.</td></tr><tr><td>Server tools</td><td><a href="./../../examples/approval-buddy/agent/tools/"><code>agent/tools/</code></a></td><td>Prepare evidence, approve, list buddies, and search GIFs.</td></tr><tr><td>Deterministic policy</td><td><a href="../../examples/approval-buddy/agent/lib/approve.ts"><code>agent/lib/approve.ts</code></a>, <a href="../../examples/approval-buddy/agent/lib/buddies.ts"><code>agent/lib/buddies.ts</code></a></td><td>Own the roster and live eligibility checks.</td></tr><tr><td>Review subagents</td><td><a href="./../../examples/approval-buddy/agent/subagents/"><code>agent/subagents/</code></a></td><td>Run deep audit and code-quality passes over the same evidence.</td></tr><tr><td>Storage</td><td><a href="../../examples/approval-buddy/agent/storage.ts"><code>agent/storage.ts</code></a></td><td>Persist sessions and events with <code>cursorHostedStorage</code> (Bugbot <code>agent_serve_*</code>).</td></tr><tr><td>Evals and unit tests</td><td><a href="./../../examples/approval-buddy/evals/"><code>evals/</code></a>, <a href="./../../examples/approval-buddy/agent/lib/"><code>agent/lib/</code></a></td><td>Protect routing, output contracts, policy, and GitHub behavior.</td></tr></tbody></table><p>There are no authored skills, MCP connections, schedules, reminders, hooks, A/B experiments, sandbox seeds, or tool approvals.</p><h2 id="prepare-credentials" tabindex="-1">Prepare credentials <a class="header-anchor" href="#prepare-credentials" aria-label="Permalink to "Prepare credentials""></a></h2><p>You need:</p><ul><li>Node 22.13 or newer.</li><li>An agent-runtime credential.</li><li>GitHub access to read PRs, post reviews, create commit statuses, and post the self-approval visibility comment.</li></ul><p>Optional GIF selection uses:</p><ul><li><code>GIPHY_API_KEY</code> or <code>APPROVAL_BUDDY_GIPHY_API_KEY</code>,</li><li><code>APPROVAL_BUDDY_STAMP_GIF</code>, or</li><li>severity-specific <code>APPROVAL_BUDDY_STAMP_GIF_<LEVEL></code> variables.</li></ul><p>If you enable Giphy in a hosted copy, declare its secret and <code>api.giphy.com</code> egress.</p><h2 id="validate-without-approving-a-pr" tabindex="-1">Validate without approving a PR <a class="header-anchor" href="#validate-without-approving-a-pr" aria-label="Permalink to "Validate without approving a PR""></a></h2><div class="language-bash vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">bash</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> validate</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dir</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> examples/approval-buddy</span></span>
|
|
22
22
|
<span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> info</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dir</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> examples/approval-buddy</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --json</span></span></code></pre></div><p>List the deterministic roster:</p><div class="language-bash vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">bash</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> call</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> list_buddies</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \</span></span>
|
|
23
23
|
<span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dir</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> examples/approval-buddy</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \</span></span>
|
|
24
24
|
<span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --input</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> '{}'</span></span></code></pre></div><p>Set a known merged PR, then run the read-only precheck:</p><div class="language-bash vp-adaptive-theme"><button title="Copy Code" class="copy"></button><span class="lang">bash</span><pre class="shiki shiki-themes github-light github-dark vp-code" tabindex="0"><code><span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">MERGED_PR_URL</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">=</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">https://github.com/your-org/your-repo/pull/123</span></span>
|
|
@@ -28,7 +28,7 @@
|
|
|
28
28
|
<span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --dir</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> examples/approval-buddy</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \</span></span>
|
|
29
29
|
<span class="line"><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> qualify/merged-pr-question</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \</span></span>
|
|
30
30
|
<span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --json</span></span></code></pre></div><p>The qualification case reads a live merged PR. The seeded format case shown by <code>--list</code> uses a planted auth-bypass diff and checks for a <code>task</code> call, both headings, and severity tags. It doesn't prove both named subagents ran or whether their output reached <code>approve_pr</code>. Unit tests under <code>agent/lib/</code> cover policy, self-approval handling, evidence limits, status mapping, GIF selection, and severity parsing.</p><h2 id="reuse-the-policy-boundary" tabindex="-1">Reuse the policy boundary <a class="header-anchor" href="#reuse-the-policy-boundary" aria-label="Permalink to "Reuse the policy boundary""></a></h2><p>Keep these properties when you replace the buddy policy:</p><ol><li>Put authorization in typed code.</li><li>Fetch the source of truth inside both prepare and mutate steps.</li><li>Give the model evidence only after the request qualifies.</li><li>Treat specialist findings as data, not authority.</li><li>Keep the core domain mutation in one named tool. Treat channel status and visibility writes as separate, audited effects.</li><li>Add a human approval gate if your policy still needs operator consent.</li><li>Test read-only routing separately from mutation.</li></ol><h2 id="where-to-go-next" tabindex="-1">Where to go next <a class="header-anchor" href="#where-to-go-next" aria-label="Permalink to "Where to go next""></a></h2><ul><li><a href="./../guides/github.html">GitHub</a></li><li><a href="./../reference/tools.html">Tools</a></li><li><a href="./../reference/subagents.html">Subagents</a></li><li><a href="./../storage.html">Storage</a></li><li><a href="./../guides/slack.html">Slack</a></li><li><a href="./../evals.html">Evals</a></li></ul></div></div></main><footer class="VPDocFooter" data-v-53b9fda6 data-v-3958cc3f><!--[--><!--]--><!----><nav class="prev-next" aria-labelledby="doc-footer-aria-label" data-v-3958cc3f><span class="visually-hidden" id="doc-footer-aria-label" data-v-3958cc3f>Pager</span><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link prev" href="/docs/example-agents/bugbot.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Previous page</span><span class="title" data-v-3958cc3f>PR evidence reviewer</span><!--]--></a></div><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link next" href="/docs/example-agents/security-reviewer.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Next page</span><span class="title" data-v-3958cc3f>Security Reviewer</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
|
|
31
|
-
<script>window.__VP_HASH_MAP__=JSON.parse("{\"ab.md\":\"BMCZ6Hd7\",\"building-with-agents.md\":\"CJCtZCyi\",\"concepts.md\":\"Cfb9b-k1\",\"deployment.md\":\"
|
|
31
|
+
<script>window.__VP_HASH_MAP__=JSON.parse("{\"ab.md\":\"BMCZ6Hd7\",\"building-with-agents.md\":\"CJCtZCyi\",\"concepts.md\":\"Cfb9b-k1\",\"deployment.md\":\"TecHo0_2\",\"evals.md\":\"DAgEc_hL\",\"example-agents_approval-buddy.md\":\"DFGBYLcc\",\"example-agents_benny.md\":\"B0gjhI-p\",\"example-agents_bugbot.md\":\"DelIdhxB\",\"example-agents_codebase-wiki.md\":\"DC6sgwn0\",\"example-agents_codeowners-review.md\":\"Ku_tG2RY\",\"example-agents_concierge.md\":\"4rQTSMXt\",\"example-agents_fsd.md\":\"CzgUrDfi\",\"example-agents_index.md\":\"BPXjUtRB\",\"example-agents_knowledge-base.md\":\"BPJiVueF\",\"example-agents_oncall.md\":\"BG_sUMly\",\"example-agents_security-reviewer.md\":\"CE9v-aKm\",\"example-agents_slack-agent.md\":\"buLbgvBf\",\"example-agents_weather-agent.md\":\"C9Qv-W0o\",\"guides_agent-to-agent.md\":\"Bpzgq2Pq\",\"guides_cloud-runtime.md\":\"gVzabdQL\",\"guides_github.md\":\"DOOCpqsW\",\"guides_human-in-the-loop.md\":\"DlUqsp1S\",\"guides_mcp-oauth.md\":\"DqmEOQY1\",\"guides_slack.md\":\"FWale6Ip\",\"guides_webhooks.md\":\"B1EswtUu\",\"hillclimbing.md\":\"DHNast08\",\"index.md\":\"t0TM2Qzz\",\"quickstart.md\":\"BU6Iwi_9\",\"reference_agent-config.md\":\"DrW2JUM8\",\"reference_channels.md\":\"DdmiKgqf\",\"reference_cli.md\":\"B74NrVzb\",\"reference_connections.md\":\"CsSr3rDz\",\"reference_hooks.md\":\"DyLVfE1O\",\"reference_http-api.md\":\"Dx_nmDG6\",\"reference_instructions.md\":\"CgoV-YEb\",\"reference_playground.md\":\"CPZhfYaO\",\"reference_project-layout.md\":\"Bd_CKtNS\",\"reference_prompt.md\":\"BaiweQxE\",\"reference_schedules.md\":\"w_F2mXB6\",\"reference_sessions.md\":\"DLd6mvbv\",\"reference_skills.md\":\"B_jHN7JL\",\"reference_subagents.md\":\"zWAMNfi1\",\"reference_tools.md\":\"BRSDnTbN\",\"scaffolding-agents.md\":\"C3pTrmoE\",\"storage.md\":\"DRTdnFvd\",\"troubleshooting.md\":\"CmQkmnzC\"}");window.__VP_SITE_DATA__=JSON.parse("{\"lang\":\"en-US\",\"dir\":\"ltr\",\"title\":\"agentkit\",\"description\":\"Filesystem-first framework for building and serving Cursor agents.\",\"base\":\"/docs/\",\"head\":[],\"router\":{\"prefetchLinks\":true},\"appearance\":true,\"themeConfig\":{\"nav\":[{\"text\":\"Quickstart\",\"link\":\"/quickstart\"},{\"text\":\"Guides\",\"link\":\"/guides/webhooks\"},{\"text\":\"Examples\",\"link\":\"/example-agents/\"},{\"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\":\"Concepts\",\"link\":\"/concepts\"}]},{\"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\":\"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\":\"Cloud runtime\",\"link\":\"/guides/cloud-runtime\"}]},{\"text\":\"Example agents\",\"items\":[{\"text\":\"Choose an example\",\"link\":\"/example-agents/\"},{\"text\":\"Weather agent\",\"link\":\"/example-agents/weather-agent\"},{\"text\":\"Slack agent\",\"link\":\"/example-agents/slack-agent\"},{\"text\":\"Concierge\",\"link\":\"/example-agents/concierge\"},{\"text\":\"Playbook router\",\"link\":\"/example-agents/benny\"},{\"text\":\"Alert investigator\",\"link\":\"/example-agents/oncall\"},{\"text\":\"PR evidence reviewer\",\"link\":\"/example-agents/bugbot\"},{\"text\":\"Approval Buddy\",\"link\":\"/example-agents/approval-buddy\"},{\"text\":\"Security Reviewer\",\"link\":\"/example-agents/security-reviewer\"},{\"text\":\"Remote PR coordinator\",\"link\":\"/example-agents/fsd\"},{\"text\":\"Knowledge base\",\"link\":\"/example-agents/knowledge-base\"},{\"text\":\"Codebase wiki\",\"link\":\"/example-agents/codebase-wiki\"},{\"text\":\"Codeowners review\",\"link\":\"/example-agents/codeowners-review\"}]},{\"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\":\"Channels\",\"link\":\"/reference/channels\"},{\"text\":\"Schedules & reminders\",\"link\":\"/reference/schedules\"},{\"text\":\"Hooks\",\"link\":\"/reference/hooks\"},{\"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>
|
|
32
32
|
|
|
33
33
|
</body>
|
|
34
34
|
</html>
|