@cursor/july 0.1.11 → 0.1.14

package/dist/docs/assets/{reference_http-api.md.Dx_nmDG6.js → reference_http-api.md.BncLd3PZ.js}

@@ -1,4 +1,4 @@
-import{_ as t,c as s,o as a,ag as o}from"./chunks/framework.CAZyNGu9.js";const u=JSON.parse('{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch.","frontmatter":{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch."},"headers":[],"relativePath":"reference/http-api.md","filePath":"reference/http-api.md"}'),n={name:"reference/http-api.md"};function i(d,e,r,l,h,c){return a(),s("div",null,[...e[0]||(e[0]=[o(`<h1 id="http-api-reference" tabindex="-1">HTTP API reference <a class="header-anchor" href="#http-api-reference" aria-label="Permalink to &quot;HTTP API reference&quot;">​</a></h1><p>Every agentkit host speaks the same stable HTTP API. In the default multi-agent layout each agent is namespaced under its slug (<code>/&lt;slug&gt;/v1/session</code>, <code>/&lt;slug&gt;/playground</code>), with host-level routes at the root. With <code>--mode single</code>, one agent serves the same surface unslugged (<code>/v1/*</code>).</p><p>Unless noted otherwise, routes run the agent&#39;s HTTP auth chain: the default is <code>localDevStrict()</code> (loopback only), replaced by <code>bearerAuth</code> under <code>--bearer-token</code> or <code>allowAll()</code> under <code>--allow-anonymous</code>. Session routes also require the caller to be the session&#39;s owner (<code>403</code> otherwise). Errors return JSON <code>{ ok: false, error: &quot;&lt;code&gt;&quot;, message? }</code> with a matching HTTP status.</p><h2 id="host-level-routes-multi-agent-mode" tabindex="-1">Host-level routes (multi-agent mode) <a class="header-anchor" href="#host-level-routes-multi-agent-mode" aria-label="Permalink to &quot;Host-level routes (multi-agent mode)&quot;">​</a></h2><p>These routes live at the host root, above any agent. The two index routes exist only while the playground is enabled (<code>--no-playground</code> removes them) and run no auth. The documentation site is mounted in both layouts and removed by <code>--no-docs</code>.</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>GET /</code></td><td>A web index of every mounted agent, linking to playgrounds (playground only)</td></tr><tr><td><code>GET /v1/agents</code></td><td>The JSON index of mounted agents (playground only, no auth)</td></tr><tr><td><code>GET /docs</code>, <code>GET /docs/*</code></td><td>This documentation, served as a static site (both layouts, no auth)</td></tr><tr><td><code>GET /v1/health</code></td><td>Host-level liveness, no auth; made for ALB/ECS checks</td></tr><tr><td><code>POST /v1/webhooks/github</code></td><td>Loopback-only trigger endpoint that fans a GitHub-shaped payload out to every mounted GitHub channel (used by local tooling)</td></tr></tbody></table><h2 id="start-a-session" tabindex="-1">Start a session <a class="header-anchor" href="#start-a-session" aria-label="Permalink to &quot;Start a session&quot;">​</a></h2><p><code>POST /v1/session</code> opens a durable conversation.</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;">curl</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> -X</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> POST</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> http://127.0.0.1:3000/</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">&lt;</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">slu</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">g</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">&gt;</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">/v1/session</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \\</span></span>
+import{_ as t,c as s,o as a,ag as o}from"./chunks/framework.CAZyNGu9.js";const u=JSON.parse('{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch.","frontmatter":{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch."},"headers":[],"relativePath":"reference/http-api.md","filePath":"reference/http-api.md"}'),n={name:"reference/http-api.md"};function d(i,e,r,l,h,c){return a(),s("div",null,[...e[0]||(e[0]=[o(`<h1 id="http-api-reference" tabindex="-1">HTTP API reference <a class="header-anchor" href="#http-api-reference" aria-label="Permalink to &quot;HTTP API reference&quot;">​</a></h1><p>Every agentkit host speaks the same stable HTTP API. In the default multi-agent layout each agent is namespaced under its slug (<code>/&lt;slug&gt;/v1/session</code>, <code>/&lt;slug&gt;/playground</code>), with host-level routes at the root. With <code>--mode single</code>, one agent serves the same surface unslugged (<code>/v1/*</code>).</p><p>Unless noted otherwise, routes run the agent&#39;s HTTP auth chain: the default is <code>localDevStrict()</code> (loopback only), replaced by <code>bearerAuth</code> under <code>--bearer-token</code> or <code>allowAll()</code> under <code>--allow-anonymous</code>. Session routes also require the caller to be the session&#39;s owner (<code>403</code> otherwise). Errors return JSON <code>{ ok: false, error: &quot;&lt;code&gt;&quot;, message? }</code> with a matching HTTP status.</p><h2 id="host-level-routes-multi-agent-mode" tabindex="-1">Host-level routes (multi-agent mode) <a class="header-anchor" href="#host-level-routes-multi-agent-mode" aria-label="Permalink to &quot;Host-level routes (multi-agent mode)&quot;">​</a></h2><p>These routes live at the host root, above any agent. The two index routes exist only while the playground is enabled (<code>--no-playground</code> removes them) and run no auth. The documentation site is mounted in both layouts and removed by <code>--no-docs</code>.</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>GET /</code></td><td>A web index of every mounted agent, linking to playgrounds (playground only)</td></tr><tr><td><code>GET /v1/agents</code></td><td>The JSON index of mounted agents (playground only, no auth)</td></tr><tr><td><code>GET /docs</code>, <code>GET /docs/*</code></td><td>This documentation, served as a static site (both layouts, no auth)</td></tr><tr><td><code>GET /v1/health</code></td><td>Host-level liveness, no auth; made for ALB/ECS checks</td></tr><tr><td><code>POST /v1/webhooks/github</code></td><td>Loopback-only trigger endpoint that fans a GitHub-shaped payload out to every mounted GitHub channel (used by local tooling)</td></tr></tbody></table><h2 id="start-a-session" tabindex="-1">Start a session <a class="header-anchor" href="#start-a-session" aria-label="Permalink to &quot;Start a session&quot;">​</a></h2><p><code>POST /v1/session</code> opens a durable conversation.</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;">curl</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> -X</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> POST</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> http://127.0.0.1:3000/</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">&lt;</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">slu</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">g</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">&gt;</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">/v1/session</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \\</span></span>
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">  -H</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &#39;content-type: application/json&#39;</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \\</span></span>
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">  -d</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &#39;{&quot;message&quot;:&quot;What can you do?&quot;}&#39;</span></span>
 <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># {&quot;ok&quot;:true,&quot;sessionId&quot;:&quot;ses_…&quot;,&quot;continuationToken&quot;:&quot;http:…&quot;,</span></span>
@@ -8,4 +8,4 @@ import{_ as t,c as s,o as a,ag as o}from"./chunks/framework.CAZyNGu9.js";const u
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">  -H</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &#39;content-type: application/json&#39;</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \\</span></span>
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">  -d</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &#39;{&quot;input&quot;:{&quot;prUrl&quot;:&quot;https://github.com/acme/checkout/pull/42&quot;}}&#39;</span></span>
 <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># {&quot;ok&quot;:true,&quot;toolName&quot;:&quot;inspect_pr&quot;,&quot;callId&quot;:&quot;tool_inspect_pr_…&quot;,</span></span>
-<span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">#  &quot;isError&quot;:false,&quot;result&quot;:{…},&quot;durationMs&quot;:12}</span></span></code></pre></div><p>It runs an authored server tool in-process: schema-validated, no model turn. An optional <code>&quot;sessionId&quot;</code> in the body runs it inside an existing session and records it on that session&#39;s stream (<code>409 session_busy</code> while a turn runs). Agent-execution tools are rejected with <code>400</code>, and unknown tools with <code>404</code> and the list of available names. For the semantics, see <a href="./tools.html#call-a-tool-without-a-model-turn">Tools</a>.</p><h2 id="discovery-and-meta" tabindex="-1">Discovery and meta <a class="header-anchor" href="#discovery-and-meta" aria-label="Permalink to &quot;Discovery and meta&quot;">​</a></h2><p>Five read-only routes describe the running agent.</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>GET /v1/info</code></td><td>The manifest snapshot: model, tools, skills, MCP connections, subagents, channels and routes (with schemas), schedules, hooks, A/B experiments, diagnostics; the same shape as <code>agentkit info --json</code></td></tr><tr><td><code>GET /v1/health</code></td><td>Per-agent liveness, no auth</td></tr><tr><td><code>GET /v1/meta</code></td><td>SPA bootstrap: agent name, dev flag, base path (no auth)</td></tr><tr><td><code>GET /v1/logs?after=N</code></td><td>Recent server log lines from the ring buffer, with a polling cursor</td></tr><tr><td><code>GET /v1/abs</code></td><td><a href="./../ab.html">Live A/B metrics</a>: per-session assignments and aggregate arm totals folded from durable event streams (<code>config</code> reports <code>maxPlaygroundSessions</code> / <code>durableSamples</code> / <code>durableSnapshots</code> from <code>agent/ab.config.ts</code>)</td></tr></tbody></table><h2 id="custom-channel-routes" tabindex="-1">Custom channel routes <a class="header-anchor" href="#custom-channel-routes" aria-label="Permalink to &quot;Custom channel routes&quot;">​</a></h2><p>Authored routes mount under <code>/v1/channels/&lt;id&gt;</code> with the methods, paths, and Zod schemas the channel declared (a <code>POST /&lt;slug&gt;/v1/channels/drive</code> route, say). Bodies are validated before handlers run (<code>400</code> on schema violations), and each channel&#39;s auth chain applies. The GitHub channel verifies <code>X-Hub-Signature-256</code> when a secret is configured. See <a href="./channels.html">Channels</a>.</p><h2 id="mcp-endpoint" tabindex="-1">MCP endpoint <a class="header-anchor" href="#mcp-endpoint" aria-label="Permalink to &quot;MCP endpoint&quot;">​</a></h2><p><code>/v1/mcp</code> serves the Model Context Protocol over streamable HTTP (stateless; POST carries the protocol, and GET/DELETE return spec-compliant 405s). The tools are <code>ask</code> (delegate a message, bounded waits), <code>check</code> (poll a running session), and <code>call_tool</code> (deterministic server-tool passthrough, present when the agent has server tools). The route runs the same auth chain as the session API. See <a href="./../guides/agent-to-agent.html">Agent-to-agent</a>.</p><h2 id="playground-eval-routes" tabindex="-1">Playground eval routes <a class="header-anchor" href="#playground-eval-routes" aria-label="Permalink to &quot;Playground eval routes&quot;">​</a></h2><p>Always registered (including production / non-<code>--dev</code> serves). The playground Evals tab uses these:</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>GET /v1/dev/evals</code></td><td>List discovered eval datapoints and project config as <code>{ evals, config }</code> (<code>config</code> includes <code>maxPlaygroundRuns</code>, <code>durableRuns</code>)</td></tr><tr><td><code>GET /v1/dev/evals/runs</code></td><td>List recent run snapshots (newest first) as <code>{ runs, activeRunId? }</code> for playground rehydrate</td></tr><tr><td><code>POST /v1/dev/evals/runs</code></td><td>Start an eval run (<code>{filterIds?, tags?}</code>); <code>202</code> with a snapshot, <code>404</code> when nothing matches, <code>409</code> when one is running</td></tr><tr><td><code>GET /v1/dev/evals/runs/:runId</code></td><td>Poll a run&#39;s progress</td></tr></tbody></table><p>Eval runs are asynchronous. Poll the run route for case progress and the final <code>completed</code> or <code>failed</code> status. Batch errors appear on the snapshot returned by the poll. Entries within <code>filterIds</code> and <code>tags</code> use OR semantics. When both fields are present, a case must match one entry from each field. Without <code>persistRuns</code> in <code>evals.config.ts</code>, listed runs are process-memory only (capped by <code>maxPlaygroundRuns</code>).</p><h2 id="dev-mode-routes" tabindex="-1">Dev-mode routes <a class="header-anchor" href="#dev-mode-routes" aria-label="Permalink to &quot;Dev-mode routes&quot;">​</a></h2><p>These routes exist only under <code>serve --dev</code>.</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>POST /v1/dev/schedules/:scheduleId</code></td><td>Dispatch a schedule by hand, exactly once, through the production path. Returns <code>{scheduleId, sessionIds}</code></td></tr><tr><td><code>GET /v1/dev/reminders</code></td><td>List reminders</td></tr><tr><td><code>POST /v1/dev/reminders/:reminderId</code></td><td>Fire a reminder by hand</td></tr></tbody></table><p>Schedules and reminders never fire automatically in dev mode. These routes are the only way they run, which keeps iteration deterministic.</p><h2 id="playground-assets" tabindex="-1">Playground assets <a class="header-anchor" href="#playground-assets" aria-label="Permalink to &quot;Playground assets&quot;">​</a></h2><p><code>GET /playground</code> and <code>GET /playground/assets/:file</code> serve the static SPA bundle (omitted with <code>--no-playground</code>). The playground calls the JSON API above and has no privileged surface.</p><h2 id="status-codes" tabindex="-1">Status codes <a class="header-anchor" href="#status-codes" aria-label="Permalink to &quot;Status codes&quot;">​</a></h2><p>Error responses use a small, consistent set of status codes.</p><table tabindex="0"><thead><tr><th>Code</th><th>Meaning here</th></tr></thead><tbody><tr><td><code>400</code></td><td>Schema-invalid body or query, agent-execution tool called on the host, malformed request</td></tr><tr><td><code>401</code></td><td>No auth policy admitted the request</td></tr><tr><td><code>403</code></td><td>Authenticated, but not the session owner</td></tr><tr><td><code>404</code></td><td>Unknown session, tool, schedule, reminder, or eval run; no eval datapoints match a run request</td></tr><tr><td><code>405</code></td><td>Wrong method (GET on the MCP endpoint, say)</td></tr><tr><td><code>409</code></td><td>Stale continuation token, busy session (<code>session_busy</code>), a non-followable task session, or an eval run already in progress</td></tr><tr><td><code>202</code></td><td>Accepted for background work (GitHub <code>{ task }</code> hooks, eval runs)</td></tr></tbody></table><h2 id="what-s-next" tabindex="-1">What&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./sessions.html">Sessions and streaming</a>: the handles and events these routes traffic in</li><li><a href="./channels.html">Channels</a>: authoring your own routes</li><li><a href="./../deployment.html">Deployment</a>: auth on real hosts</li></ul>`,53)])])}const k=t(n,[["render",i]]);export{u as __pageData,k as default};
+<span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">#  &quot;isError&quot;:false,&quot;result&quot;:{…},&quot;durationMs&quot;:12}</span></span></code></pre></div><p>It runs an authored server tool in-process: schema-validated, no model turn. An optional <code>&quot;sessionId&quot;</code> in the body runs it inside an existing session and records it on that session&#39;s stream (<code>409 session_busy</code> while a turn runs). Agent-execution tools are rejected with <code>400</code>, and unknown tools with <code>404</code> and the list of available names. For the semantics, see <a href="./tools.html#call-a-tool-without-a-model-turn">Tools</a>.</p><h2 id="discovery-and-meta" tabindex="-1">Discovery and meta <a class="header-anchor" href="#discovery-and-meta" aria-label="Permalink to &quot;Discovery and meta&quot;">​</a></h2><p>Five read-only routes describe the running agent.</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>GET /v1/info</code></td><td>The manifest snapshot: model, tools, skills, MCP connections, subagents, channels and routes (with schemas), schedules, hooks, A/B experiments, diagnostics; the same shape as <code>agentkit info --json</code></td></tr><tr><td><code>GET /v1/health</code></td><td>Per-agent liveness, no auth</td></tr><tr><td><code>GET /v1/meta</code></td><td>SPA bootstrap: agent name, dev flag, base path (no auth)</td></tr><tr><td><code>GET /v1/logs?after=N</code></td><td>Recent server log lines from the ring buffer, with a polling cursor</td></tr><tr><td><code>GET /v1/abs</code></td><td><a href="./../ab.html">Live A/B metrics</a>: per-session assignments and aggregate arm totals folded from durable event streams (<code>config</code> reports <code>maxPlaygroundSessions</code> / <code>durableSamples</code> / <code>durableSnapshots</code> from <code>agent/ab.config.ts</code>)</td></tr></tbody></table><h2 id="custom-channel-routes" tabindex="-1">Custom channel routes <a class="header-anchor" href="#custom-channel-routes" aria-label="Permalink to &quot;Custom channel routes&quot;">​</a></h2><p>Authored routes mount under <code>/v1/channels/&lt;id&gt;</code> with the methods, paths, and Zod schemas the channel declared (a <code>POST /&lt;slug&gt;/v1/channels/drive</code> route, say). Bodies are validated before handlers run (<code>400</code> on schema violations), and each channel&#39;s auth chain applies. The GitHub channel verifies <code>X-Hub-Signature-256</code> when a secret is configured. See <a href="./channels.html">Channels</a>.</p><h2 id="mcp-endpoint" tabindex="-1">MCP endpoint <a class="header-anchor" href="#mcp-endpoint" aria-label="Permalink to &quot;MCP endpoint&quot;">​</a></h2><p><code>/v1/mcp</code> serves the Model Context Protocol over streamable HTTP (stateless; POST carries the protocol, and GET/DELETE return spec-compliant 405s). The tools are <code>ask</code> (delegate a message, bounded waits), <code>check</code> (poll a running session), and <code>call_tool</code> (deterministic server-tool passthrough, present when the agent has server tools). The route runs the same auth chain as the session API. See <a href="./../guides/agent-to-agent.html">Agent-to-agent</a>.</p><h2 id="playground-eval-routes" tabindex="-1">Playground eval routes <a class="header-anchor" href="#playground-eval-routes" aria-label="Permalink to &quot;Playground eval routes&quot;">​</a></h2><p>Always registered (including production / non-<code>--dev</code> serves). The playground Evals tab uses these:</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>GET /v1/dev/evals</code></td><td>List discovered eval datapoints and project config as <code>{ evals, config }</code> (<code>config</code> includes <code>maxPlaygroundRuns</code>, <code>durableRuns</code>)</td></tr><tr><td><code>GET /v1/dev/evals/runs</code></td><td>List recent run snapshots (newest first) as <code>{ runs, activeRunId? }</code> for playground rehydrate</td></tr><tr><td><code>POST /v1/dev/evals/runs</code></td><td>Start an eval run (<code>{filterIds?, tags?}</code>); <code>202</code> with a snapshot (<code>runId</code> is the Eval ID), <code>404</code> when nothing matches, <code>409</code> when one is running</td></tr><tr><td><code>GET /v1/dev/evals/runs/:runId</code></td><td>Poll a run&#39;s progress</td></tr><tr><td><code>POST /v1/dev/evals/runs/:runId/cancel</code></td><td>Cancel a running batch; <code>200</code> with snapshot, <code>404</code> unknown, <code>409</code> when not running</td></tr></tbody></table><p>Eval runs are asynchronous. Poll the run route for case progress and the final <code>completed</code> or <code>failed</code> status. Batch errors appear on the snapshot returned by the poll. Entries within <code>filterIds</code> and <code>tags</code> use OR semantics. When both fields are present, a case must match one entry from each field. Without <code>persistRuns</code> in <code>evals.config.ts</code>, listed runs are process-memory only (capped by <code>maxPlaygroundRuns</code>).</p><h2 id="dev-mode-routes" tabindex="-1">Dev-mode routes <a class="header-anchor" href="#dev-mode-routes" aria-label="Permalink to &quot;Dev-mode routes&quot;">​</a></h2><p>These routes exist only under <code>serve --dev</code>.</p><table tabindex="0"><thead><tr><th>Route</th><th>What it does</th></tr></thead><tbody><tr><td><code>POST /v1/dev/schedules/:scheduleId</code></td><td>Dispatch a schedule by hand, exactly once, through the production path. Returns <code>{scheduleId, sessionIds}</code></td></tr><tr><td><code>GET /v1/dev/reminders</code></td><td>List reminders</td></tr><tr><td><code>POST /v1/dev/reminders/:reminderId</code></td><td>Fire a reminder by hand</td></tr></tbody></table><p>Schedules and reminders never fire automatically in dev mode. These routes are the only way they run, which keeps iteration deterministic.</p><h2 id="playground-assets" tabindex="-1">Playground assets <a class="header-anchor" href="#playground-assets" aria-label="Permalink to &quot;Playground assets&quot;">​</a></h2><p><code>GET /playground</code> and <code>GET /playground/assets/:file</code> serve the static SPA bundle (omitted with <code>--no-playground</code>). The playground calls the JSON API above and has no privileged surface.</p><h2 id="status-codes" tabindex="-1">Status codes <a class="header-anchor" href="#status-codes" aria-label="Permalink to &quot;Status codes&quot;">​</a></h2><p>Error responses use a small, consistent set of status codes.</p><table tabindex="0"><thead><tr><th>Code</th><th>Meaning here</th></tr></thead><tbody><tr><td><code>400</code></td><td>Schema-invalid body or query, agent-execution tool called on the host, malformed request</td></tr><tr><td><code>401</code></td><td>No auth policy admitted the request</td></tr><tr><td><code>403</code></td><td>Authenticated, but not the session owner</td></tr><tr><td><code>404</code></td><td>Unknown session, tool, schedule, reminder, or eval run; no eval datapoints match a run request</td></tr><tr><td><code>405</code></td><td>Wrong method (GET on the MCP endpoint, say)</td></tr><tr><td><code>409</code></td><td>Stale continuation token, busy session (<code>session_busy</code>), a non-followable task session, or an eval run already in progress</td></tr><tr><td><code>202</code></td><td>Accepted for background work (GitHub <code>{ task }</code> hooks, eval runs)</td></tr></tbody></table><h2 id="what-s-next" tabindex="-1">What&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./sessions.html">Sessions and streaming</a>: the handles and events these routes traffic in</li><li><a href="./channels.html">Channels</a>: authoring your own routes</li><li><a href="./../deployment.html">Deployment</a>: auth on real hosts</li></ul>`,53)])])}const k=t(n,[["render",d]]);export{u as __pageData,k as default};

package/dist/docs/assets/{reference_http-api.md.Dx_nmDG6.lean.js → reference_http-api.md.BncLd3PZ.lean.js}

@@ -1 +1 @@
-import{_ as t,c as s,o as a,ag as o}from"./chunks/framework.CAZyNGu9.js";const u=JSON.parse('{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch.","frontmatter":{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch."},"headers":[],"relativePath":"reference/http-api.md","filePath":"reference/http-api.md"}'),n={name:"reference/http-api.md"};function i(d,e,r,l,h,c){return a(),s("div",null,[...e[0]||(e[0]=[o("",53)])])}const k=t(n,[["render",i]]);export{u as __pageData,k as default};
+import{_ as t,c as s,o as a,ag as o}from"./chunks/framework.CAZyNGu9.js";const u=JSON.parse('{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch.","frontmatter":{"title":"HTTP API","description":"Every route the server mounts: sessions, approvals, deterministic tool calls, discovery, the MCP endpoint, and dev-mode dispatch."},"headers":[],"relativePath":"reference/http-api.md","filePath":"reference/http-api.md"}'),n={name:"reference/http-api.md"};function d(i,e,r,l,h,c){return a(),s("div",null,[...e[0]||(e[0]=[o("",53)])])}const k=t(n,[["render",d]]);export{u as __pageData,k as default};

package/dist/docs/building-with-agents.html

@@ -9,9 +9,9 @@
     <link rel="preload stylesheet" href="/docs/assets/style.Bj-IXa5W.css" as="style">
     <link rel="preload stylesheet" href="/docs/vp-icons.css" as="style">
     
-    <script type="module" src="/docs/assets/app.BcXUGca7.js"></script>
+    <script type="module" src="/docs/assets/app.CKFFLL4Z.js"></script>
     <link rel="preload" href="/docs/assets/inter-roman-latin.Di8DUHzh.woff2" as="font" type="font/woff2" crossorigin="">
-    <link rel="modulepreload" href="/docs/assets/chunks/theme.BnvulOX2.js">
+    <link rel="modulepreload" href="/docs/assets/chunks/theme.BXSPBMyn.js">
     <link rel="modulepreload" href="/docs/assets/chunks/framework.CAZyNGu9.js">
     <link rel="modulepreload" href="/docs/assets/building-with-agents.md.CJCtZCyi.lean.js">
     <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>
@@ -31,7 +31,7 @@
 <span class="line"></span>
 <span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</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;"> --list</span></span>
 <span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</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;"> --json</span></span></code></pre></div><p><code>serve</code> boots without an API key, which is enough to check channel mounts and the playground shell. Model turns need a credential. When the environment lacks one, finish every key-free check, then hand the <code>run</code> and <code>eval</code> commands to the operator.</p><p>Test server tools with <code>call</code> before tuning the prompt. It runs a tool in-process with schema validation and no model turn. If the tool returns the wrong data, a prompt change won&#39;t fix it.</p><p><code>validate</code> and <code>run</code> don&#39;t type-check the project because tsx strips types. Run the project&#39;s TypeScript check before shipping. Tool results must also be JSON-shaped. Use object literals or <code>type</code> aliases for <code>execute</code> return types instead of <code>interface</code> types.</p><h2 id="how-do-i-improve-the-agent-after-its-first-run" tabindex="-1">How do I improve the agent after its first run? <a class="header-anchor" href="#how-do-i-improve-the-agent-after-its-first-run" aria-label="Permalink to &quot;How do I improve the agent after its first run?&quot;">​</a></h2><p>Scaffolding proves the agent runs. It says nothing about quality. Once a smoke turn passes, give the hillclimb skill:</p><ol><li><strong>Fixtures</strong>: one to three fixed inputs, such as a PR URL, a canonical question, or a saved webhook payload</li><li><strong>Success criteria</strong>: correct tool choice, fewer tool calls, lower wall time, or better output</li><li><strong>The freeze line</strong>: the API contract, output shape, and existing evals that must stay unchanged</li></ol><p>Have the coding agent read <a href="./../skills/hillclimb/SKILL.html"><code>skills/hillclimb/SKILL.md</code></a>. It measures the current run, proposes one change, remeasures the same fixtures, and adds an eval for each kept improvement.</p><h2 id="related" tabindex="-1">Related <a class="header-anchor" href="#related" aria-label="Permalink to &quot;Related&quot;">​</a></h2><ul><li><a href="./quickstart.html">Create your first agent</a></li><li><a href="./scaffolding-agents.html">Scaffold an agent with Cursor</a></li><li><a href="./evals.html">Evals</a></li><li><a href="./hillclimbing.html">Hillclimbing</a></li><li><a href="./reference/cli.html">CLI reference</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/concepts.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Previous page</span><span class="title" data-v-3958cc3f>Concepts</span><!--]--></a></div><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link next" href="/docs/evals.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Next page</span><span class="title" data-v-3958cc3f>Evals</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
-    <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>
+    <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\":\"DYOjkRCX\",\"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\":\"DnYfr5V2\",\"reference_connections.md\":\"CsSr3rDz\",\"reference_hooks.md\":\"DyLVfE1O\",\"reference_http-api.md\":\"BncLd3PZ\",\"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>
     
   </body>
 </html>

package/dist/docs/concepts.html

@@ -9,9 +9,9 @@
     <link rel="preload stylesheet" href="/docs/assets/style.Bj-IXa5W.css" as="style">
     <link rel="preload stylesheet" href="/docs/vp-icons.css" as="style">
     
-    <script type="module" src="/docs/assets/app.BcXUGca7.js"></script>
+    <script type="module" src="/docs/assets/app.CKFFLL4Z.js"></script>
     <link rel="preload" href="/docs/assets/inter-roman-latin.Di8DUHzh.woff2" as="font" type="font/woff2" crossorigin="">
-    <link rel="modulepreload" href="/docs/assets/chunks/theme.BnvulOX2.js">
+    <link rel="modulepreload" href="/docs/assets/chunks/theme.BXSPBMyn.js">
     <link rel="modulepreload" href="/docs/assets/chunks/framework.CAZyNGu9.js">
     <link rel="modulepreload" href="/docs/assets/concepts.md.Cfb9b-k1.lean.js">
     <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>
@@ -22,7 +22,7 @@
 <span class="line"><span>  sessions/&lt;id&gt;/events.ndjson</span></span>
 <span class="line"><span>  sessions/&lt;id&gt;/workspace/</span></span>
 <span class="line"><span>  traces/&lt;sessionId&gt;.ndjson</span></span></code></pre></div><h2 id="how-can-one-agent-call-another" tabindex="-1">How can one agent call another? <a class="header-anchor" href="#how-can-one-agent-call-another" aria-label="Permalink to &quot;How can one agent call another?&quot;">​</a></h2><p>Every mounted agent also serves MCP at <code>/&lt;slug&gt;/v1/mcp</code>. Another agent or MCP client can use <code>ask</code>, <code>check</code>, and <code>call_tool</code> to delegate work. A peer MCP connection such as <code>defineConnection({ agent: &quot;weather-agent&quot; })</code> adds those tools to the calling agent.</p><p>See <a href="./guides/agent-to-agent.html">Agent-to-agent</a> for a complete example.</p><h2 id="which-rules-prevent-common-setup-problems" tabindex="-1">Which rules prevent common setup problems? <a class="header-anchor" href="#which-rules-prevent-common-setup-problems" aria-label="Permalink to &quot;Which rules prevent common setup problems?&quot;">​</a></h2><ul><li>Use Node 22.13 or newer. Bun isn&#39;t supported.</li><li>Put evals under the project-root <code>evals/</code> directory, not <code>agent/evals/</code>.</li><li>Run a TypeScript check before shipping. <code>validate</code> and <code>run</code> execute TypeScript but don&#39;t type-check it.</li><li>Return JSON-shaped values from tool <code>execute</code> functions.</li><li>Keep local session workspaces away from parent rules you don&#39;t want the agent to inherit.</li><li>Sign in or set <code>CURSOR_API_KEY</code> before starting a model turn. Discovery, validation, direct tool calls, and server startup work without a credential.</li></ul><h2 id="related" tabindex="-1">Related <a class="header-anchor" href="#related" aria-label="Permalink to &quot;Related&quot;">​</a></h2><ul><li><a href="./quickstart.html">Quickstart</a></li><li><a href="./reference/project-layout.html">Project layout</a></li><li><a href="./reference/sessions.html">Sessions and streaming</a></li><li><a href="./reference/channels.html">Channels</a></li><li><a href="./ab.html">Live A/B metrics</a></li><li><a href="./guides/cloud-runtime.html">Cloud runtime</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/scaffolding-agents.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Previous page</span><span class="title" data-v-3958cc3f>Scaffold an agent with Cursor</span><!--]--></a></div><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link next" href="/docs/building-with-agents.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Next page</span><span class="title" data-v-3958cc3f>Building agents with agents</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
-    <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>
+    <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\":\"DYOjkRCX\",\"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\":\"DnYfr5V2\",\"reference_connections.md\":\"CsSr3rDz\",\"reference_hooks.md\":\"DyLVfE1O\",\"reference_http-api.md\":\"BncLd3PZ\",\"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>
     
   </body>
 </html>

package/dist/docs/deployment.html

@@ -9,9 +9,9 @@
     <link rel="preload stylesheet" href="/docs/assets/style.Bj-IXa5W.css" as="style">
     <link rel="preload stylesheet" href="/docs/vp-icons.css" as="style">
     
-    <script type="module" src="/docs/assets/app.BcXUGca7.js"></script>
+    <script type="module" src="/docs/assets/app.CKFFLL4Z.js"></script>
     <link rel="preload" href="/docs/assets/inter-roman-latin.Di8DUHzh.woff2" as="font" type="font/woff2" crossorigin="">
-    <link rel="modulepreload" href="/docs/assets/chunks/theme.BnvulOX2.js">
+    <link rel="modulepreload" href="/docs/assets/chunks/theme.BXSPBMyn.js">
     <link rel="modulepreload" href="/docs/assets/chunks/framework.CAZyNGu9.js">
     <link rel="modulepreload" href="/docs/assets/deployment.md.TecHo0_2.lean.js">
     <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>
@@ -73,7 +73,7 @@
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">  --state-root</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> /var/lib/agent-serve</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> \</span></span>
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">  --bearer-token</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">$AGENTKIT_BEARER_TOKEN</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">&quot;</span></span></code></pre></div><p>Mount the state root as a persistent volume and inject secrets at startup. Install <code>git</code> and <code>gh</code> when channels need host-side GitHub work. Don&#39;t put secrets in the image.</p><h3 id="serve-many-agents-from-one-process" tabindex="-1">Serve many agents from one process <a class="header-anchor" href="#serve-many-agents-from-one-process" aria-label="Permalink to &quot;Serve many agents from one process&quot;">​</a></h3><p>Point <code>serve</code> at a folder of agent projects and every child mounts under its directory name on one port. One process, one state root, one credential:</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;"> /srv/agents</span></span>
 <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># index at /, each agent at /&lt;slug&gt;/v1/*, /&lt;slug&gt;/playground</span></span></code></pre></div><p>Only mount what you mean to run. Every mounted agent&#39;s channels are live, and webhook-driven agents spend model budget on every wake. <code>--mode single</code> serves exactly one agent at the unslugged <code>/v1/*</code> when the agent is the whole host. See the <a href="./reference/http-api.html">HTTP API</a> for route layout and the <a href="./guides/slack.html">Slack guide</a> for multi-agent token setup.</p><h3 id="the-production-flags" tabindex="-1">The production flags <a class="header-anchor" href="#the-production-flags" aria-label="Permalink to &quot;The production flags&quot;">​</a></h3><p>Use these settings in production:</p><table tabindex="0"><thead><tr><th>Flag</th><th>In production</th></tr></thead><tbody><tr><td><code>--dev</code></td><td>Leave off. Dev mode admits unsigned loopback GitHub deliveries, widens playground session listing on loopback, and never auto-fires schedules.</td></tr><tr><td><code>--bearer-token</code></td><td>Set on shared hosts unless an authenticating proxy is the trust boundary and you use <code>--allow-anonymous</code> instead.</td></tr><tr><td><code>--allow-anonymous</code></td><td>Use only behind an authenticating network boundary. It also widens playground session access so Slack and webhook sessions appear.</td></tr><tr><td><code>--state-root</code></td><td>Place on a persistent volume outside any repo.</td></tr><tr><td><code>--public-url</code></td><td>Set when cloud-runtime turns must call back into peers on this host.</td></tr><tr><td><code>--no-playground</code></td><td>Set when no human needs the UI.</td></tr><tr><td><code>--no-docs</code></td><td>Set to remove the documentation site at <code>/docs</code>.</td></tr><tr><td><code>--no-schedules</code></td><td>Set on secondary hosts so schedules run exactly once.</td></tr></tbody></table><p>Schedules fire on their cron cadence (UTC) in production mode. They have no cross-host coordination, so enable them on exactly one serving process per project.</p><h3 id="restarts-and-upgrades" tabindex="-1">Restarts and upgrades <a class="header-anchor" href="#restarts-and-upgrades" aria-label="Permalink to &quot;Restarts and upgrades&quot;">​</a></h3><p>Restarts preserve sessions, event streams, and SDK conversation state under the state root. Parked approvals and in-memory reminders don&#39;t survive a restart; re-run or recreate them afterward.</p><h3 id="observability" tabindex="-1">Observability <a class="header-anchor" href="#observability" aria-label="Permalink to &quot;Observability&quot;">​</a></h3><p>Use <a href="./reference/cli.html#logs"><code>agentkit logs</code></a> for runtime output, <a href="./reference/hooks.html">hooks</a> for metrics export, and <a href="./reference/sessions.html#how-do-i-inspect-a-saved-event-stream">session traces</a> for incident review.</p><h2 id="what-s-next" tabindex="-1">What&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./reference/cli.html#deploy">CLI reference</a>: deploy, inspect, stop, and rotate hosted agents</li><li><a href="./storage.html">Storage</a>: preserve supported records across engine replacements</li><li><a href="./reference/channels.html#auth-policies">Channels</a>: the auth policies in detail</li><li><a href="./guides/github.html">GitHub guide</a>: delivery paths without a public URL</li><li><a href="./troubleshooting.html">Troubleshooting</a>: the symptom table for when a deploy misbehaves</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/codeowners-review.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Previous page</span><span class="title" data-v-3958cc3f>Codeowners review</span><!--]--></a></div><div class="pager" data-v-3958cc3f><a class="VPLink link pager-link next" href="/docs/troubleshooting.html" data-v-3958cc3f><!--[--><span class="desc" data-v-3958cc3f>Next page</span><span class="title" data-v-3958cc3f>Fix common problems</span><!--]--></a></div></nav></footer><!--[--><!--]--></div></div></div><!--[--><!--]--></div></div><!----><!--[--><!--]--></div></div>
-    <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>
+    <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\":\"DYOjkRCX\",\"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\":\"DnYfr5V2\",\"reference_connections.md\":\"CsSr3rDz\",\"reference_hooks.md\":\"DyLVfE1O\",\"reference_http-api.md\":\"BncLd3PZ\",\"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>
     
   </body>
 </html>

package/dist/docs/evals.html

@@ -9,11 +9,11 @@
     <link rel="preload stylesheet" href="/docs/assets/style.Bj-IXa5W.css" as="style">
     <link rel="preload stylesheet" href="/docs/vp-icons.css" as="style">
     
-    <script type="module" src="/docs/assets/app.BcXUGca7.js"></script>
+    <script type="module" src="/docs/assets/app.CKFFLL4Z.js"></script>
     <link rel="preload" href="/docs/assets/inter-roman-latin.Di8DUHzh.woff2" as="font" type="font/woff2" crossorigin="">
-    <link rel="modulepreload" href="/docs/assets/chunks/theme.BnvulOX2.js">
+    <link rel="modulepreload" href="/docs/assets/chunks/theme.BXSPBMyn.js">
     <link rel="modulepreload" href="/docs/assets/chunks/framework.CAZyNGu9.js">
-    <link rel="modulepreload" href="/docs/assets/evals.md.DAgEc_hL.lean.js">
+    <link rel="modulepreload" href="/docs/assets/evals.md.DYOjkRCX.lean.js">
     <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>
     <script id="check-mac-os">document.documentElement.classList.toggle("mac",/Mac|iPhone|iPod|iPad/i.test(navigator.platform));</script>
   </head>
@@ -60,7 +60,7 @@
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">      },</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">    },</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">  ],</span></span>
-<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">});</span></span></code></pre></div><p>Case ids must be single path segments, unique within the file. Each case can set its own <code>description</code>, <code>tags</code>, and <code>timeoutMs</code>. A case-level value replaces the file-level value for that datapoint.</p><h2 id="configure-eval-runs" tabindex="-1">Configure eval runs <a class="header-anchor" href="#configure-eval-runs" aria-label="Permalink to &quot;Configure eval runs&quot;">​</a></h2><p>Each project with evals needs <code>evals/evals.config.ts</code> or <code>evals/evals.config.js</code>, and it must set <code>maxConcurrency</code>. Each case issues real model-provider requests, so concurrency is capped hard at 200. Existing projects use 20. Discovery with <code>eval --list</code> works without this file, but running a case does not.</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:#D73A49;--shiki-dark:#F97583;">import</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> {</span></span>
+<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">});</span></span></code></pre></div><p>Case ids must be single path segments, unique within the file. Each case can set its own <code>description</code>, <code>tags</code>, <code>timeoutMs</code>, and <code>iterations</code>. A case-level value replaces the file-level value for that datapoint.</p><h3 id="iterations" tabindex="-1">Iterations <a class="header-anchor" href="#iterations" aria-label="Permalink to &quot;Iterations&quot;">​</a></h3><p><code>iterations</code> (file or case, default <code>1</code>) runs a datapoint repeatedly. Discovery expands <code>iterations: 3</code> on case <code>nyc</code> to runnable ids <code>weather/nyc/1</code>, <code>weather/nyc/2</code>, <code>weather/nyc/3</code> (filter prefix <code>weather/nyc</code> still selects all three). Each expanded case exposes <code>t.iteration</code> / <code>t.iterations</code> on the test context. Cap is 100.</p><p><code>maxConcurrency</code> counts <strong>authored datapoints</strong>, not expanded iterations: siblings <code>…/1</code>…<code>…/n</code> share one concurrency slot and run sequentially. A suite with 11 cases × 3 iterations and <code>maxConcurrency: 20</code> therefore has at most 11 cases in flight, not 33.</p><h2 id="configure-eval-runs" tabindex="-1">Configure eval runs <a class="header-anchor" href="#configure-eval-runs" aria-label="Permalink to &quot;Configure eval runs&quot;">​</a></h2><p>Each project with evals needs <code>evals/evals.config.ts</code> or <code>evals/evals.config.js</code>, and it must set <code>maxConcurrency</code>. Each case issues real model-provider requests, so concurrency is capped hard at 200. Existing projects use 20. Discovery with <code>eval --list</code> works without this file, but running a case does not.</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:#D73A49;--shiki-dark:#F97583;">import</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> {</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">  defineEvalConfig,</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">  persistEvalRunsToDir,</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">} </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">from</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &quot;@cursor/july/evals&quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">;</span></span>
@@ -74,7 +74,7 @@
 <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">  // when `serve` exits. Opt into durable storage:</span></span>
 <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">  // persistRuns: persistEvalRunsToDir(&quot;.agent-serve/eval-runs&quot;),</span></span>
 <span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;">  // Or implement { load, save, delete } yourself (S3, DB, …).</span></span>
-<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">});</span></span></code></pre></div><p>The timeout order is case or file <code>timeoutMs</code>, CLI <code>--timeout-ms</code>, project config <code>timeoutMs</code>, then the 180-second runner default.</p><p>Playground batch retention is separate from case concurrency:</p><table tabindex="0"><thead><tr><th>Option</th><th>Default</th><th>Meaning</th></tr></thead><tbody><tr><td><code>maxPlaygroundRuns</code></td><td><code>20</code></td><td>Max batches in the playground / <code>/v1/dev/evals*</code> history (not CLI <code>eval</code>)</td></tr><tr><td><code>persistRuns</code></td><td>unset</td><td>Optional <code>{ load, save, delete }</code> so batches survive process restart (<code>delete</code> required for durable prune)</td></tr></tbody></table><p>Without <code>persistRuns</code>, navigating away and back still works while the same <code>serve</code> process is up; a restart clears history.</p><h2 id="drive-and-assert-with-t" tabindex="-1">Drive and assert with <code>t</code> <a class="header-anchor" href="#drive-and-assert-with-t" aria-label="Permalink to &quot;Drive and assert with `t`&quot;">​</a></h2><p><code>t</code> is both the driver and the assertion surface. You write ordinary control flow, sending turns and asserting inline.</p><p>Drive the agent with <code>t.send(message, options?)</code>. It runs one turn and waits for the session to park or fail. Multiple sends in one case share the session, which is how you write multi-turn evals. The return value contains the turn&#39;s <code>message</code>, <code>sessionId</code>, <code>events</code>, <code>toolCalls</code>, and <code>ok</code> state.</p><p>Read the full case state with <code>t.reply</code> (the last assistant text), <code>t.events</code> (every captured session event across turns), and <code>t.sessionId</code>.</p><p>Assert with the gates:</p><table tabindex="0"><thead><tr><th>Gate</th><th>Checks</th></tr></thead><tbody><tr><td><code>t.succeeded()</code></td><td>the captured trajectory has at least one turn and did not fail</td></tr><tr><td><code>t.calledTool(name)</code></td><td><code>name</code> appears in the captured tool calls</td></tr><tr><td><code>t.notCalledTool(name)</code></td><td><code>name</code> does not appear in the captured tool calls</td></tr><tr><td><code>t.messageIncludes(token)</code></td><td>the last assistant reply matches a string or <code>RegExp</code></td></tr><tr><td><code>t.check(value, expectation)</code></td><td>any value, against a builder</td></tr></tbody></table><p><code>calledTool</code> reads the recorded trajectory. A requested call counts even when its result has not arrived. To require a completed result, inspect <code>t.events</code> for an <code>action.result</code> event.</p><p>The expectation builders are <code>includes(string | RegExp)</code>, <code>equals(value)</code>, and <code>satisfies(predicate, label)</code>. <code>includes</code> stringifies its input, <code>equals</code> compares values deeply, and <code>satisfies</code> runs your predicate. <code>t.log(message)</code> records a debug line for the CLI and playground result.</p><p>Two <code>t.send</code> options make cases self-contained: <code>workspaceFiles</code> (<code>{ path: contents }</code>, seeded into the session workspace before the first turn) and <code>workspaceDir</code> (an absolute harness cwd). Prefer <code>workspaceFiles</code> over machine-local paths, so the case carries its own fixtures. Both options only apply to the first <code>t.send</code>, when the runner creates the session.</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:#D73A49;--shiki-dark:#F97583;">const</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> toolResults</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;"> =</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> t.events.</span><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">filter</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">((</span><span style="--shiki-light:#E36209;--shiki-dark:#FFAB70;">e</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">) </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">=&gt;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> e.type </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">===</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &quot;action.result&quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">);</span></span>
+<span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">});</span></span></code></pre></div><p>The timeout order is case or file <code>timeoutMs</code>, CLI <code>--timeout-ms</code>, project config <code>timeoutMs</code>, then the 180-second runner default.</p><p>Playground batch retention is separate from case concurrency:</p><table tabindex="0"><thead><tr><th>Option</th><th>Default</th><th>Meaning</th></tr></thead><tbody><tr><td><code>maxPlaygroundRuns</code></td><td><code>20</code></td><td>Max batches in the playground / <code>/v1/dev/evals*</code> history (not CLI <code>eval</code>)</td></tr><tr><td><code>persistRuns</code></td><td>unset</td><td>Optional <code>{ load, save, delete }</code> so batches survive process restart (<code>delete</code> required for durable prune)</td></tr></tbody></table><p>Without <code>persistRuns</code>, navigating away and back still works while the same <code>serve</code> process is up; a restart clears history.</p><h2 id="drive-and-assert-with-t" tabindex="-1">Drive and assert with <code>t</code> <a class="header-anchor" href="#drive-and-assert-with-t" aria-label="Permalink to &quot;Drive and assert with `t`&quot;">​</a></h2><p><code>t</code> is both the driver and the assertion surface. You write ordinary control flow, sending turns and asserting inline.</p><p>Drive the agent with <code>t.send(message, options?)</code>. It runs one turn and waits for the session to park or fail. Multiple sends in one case share the session, which is how you write multi-turn evals. The return value contains the turn&#39;s <code>message</code>, <code>sessionId</code>, <code>events</code>, <code>toolCalls</code>, and <code>ok</code> state.</p><p>Read the full case state with <code>t.reply</code> (the last assistant text), <code>t.events</code> (every captured session event across turns), and <code>t.sessionId</code>.</p><p>Assert with the gates:</p><table tabindex="0"><thead><tr><th>Gate</th><th>Checks</th></tr></thead><tbody><tr><td><code>t.succeeded()</code></td><td>the captured trajectory has at least one turn and did not fail</td></tr><tr><td><code>t.calledTool(name)</code></td><td><code>name</code> appears in the captured tool calls</td></tr><tr><td><code>t.notCalledTool(name)</code></td><td><code>name</code> does not appear in the captured tool calls</td></tr><tr><td><code>t.messageIncludes(token)</code></td><td>the last assistant reply matches a string or <code>RegExp</code></td></tr><tr><td><code>t.check(value, expectation)</code></td><td>any value, against a builder</td></tr></tbody></table><p><code>calledTool</code> reads the recorded trajectory. A requested call counts even when its result has not arrived. To require a completed result, inspect <code>t.events</code> for an <code>action.result</code> event.</p><p>The expectation builders are <code>includes(string | RegExp)</code>, <code>equals(value)</code>, and <code>satisfies(predicate, label)</code>. <code>includes</code> stringifies its input, <code>equals</code> compares values deeply, and <code>satisfies</code> runs your predicate. <code>t.log(message)</code> records a debug line for the CLI and playground result.</p><p>Three <code>t.send</code> options apply on session create (first <code>t.send</code> only):</p><ul><li><code>workspaceFiles</code> — <code>{ path: contents }</code>, seeded into the local session workspace. Prefer this over machine-local paths.</li><li><code>workspaceDir</code> — absolute harness cwd (local runtime).</li><li><code>cloud</code> — per-session cloud options merged over the agent&#39;s static <code>cloud</code> config (repos / env / …). Use a pinned <code>repos</code> override to attach a fixture repo for cloud evals without putting it on the agent&#39;s default <code>cloud.repos</code>. Cloud ignores <code>workspaceFiles</code> seeds.</li></ul><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:#D73A49;--shiki-dark:#F97583;">const</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> toolResults</span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;"> =</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> t.events.</span><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">filter</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">((</span><span style="--shiki-light:#E36209;--shiki-dark:#FFAB70;">e</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">) </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">=&gt;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> e.type </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">===</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> &quot;action.result&quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">);</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">t.</span><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">check</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">(</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">  toolResults.</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">length</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">,</span></span>
 <span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">  satisfies</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">((</span><span style="--shiki-light:#E36209;--shiki-dark:#FFAB70;">n</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">) </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">=&gt;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;"> (n </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">as</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> number</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">) </span><span style="--shiki-light:#D73A49;--shiki-dark:#F97583;">&lt;=</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> 4</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">, </span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;">&quot;at most 4 tool calls&quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">)</span></span>
@@ -102,8 +102,14 @@
 <span class="line"><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">      &quot;durationMs&quot;</span><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">: </span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;">12340</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">    }</span></span>
 <span class="line"><span style="--shiki-light:#24292E;--shiki-dark:#E1E4E8;">  ]</span></span>
-<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 &quot;Run evals in the playground&quot;">​</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 &quot;What good cases assert&quot;">​</a></h2><p>Gate decisions and shape, not prose. Model wording varies run to run. Tool choice, tool avoidance, and output structure are the stable contract.</p><ol><li><code>t.succeeded()</code>: always, first.</li><li>The tool decision: <code>calledTool</code> for the intended path, <code>notCalledTool</code> for the likely wrong alternative. The pair is stronger than either alone.</li><li>Output shape: a regex for the contract (<code>/ready|blocked/i</code>, a JSON marker, a findings-block fence), never exact sentences.</li><li>For structured output, parse <code>t.reply</code> and check fields with <code>satisfies</code> instead of substring-matching JSON.</li></ol><p>The common failure modes: asserting exact phrasing, packing more than about five gates into one case (split it), and cases that depend on live external state that drifts (pin the input; see fixtures).</p><h2 id="pick-fixtures-by-agent-type" tabindex="-1">Pick fixtures by agent type <a class="header-anchor" href="#pick-fixtures-by-agent-type" aria-label="Permalink to &quot;Pick fixtures by agent type&quot;">​</a></h2><p>The right fixture depends on the surface under test.</p><table tabindex="0"><thead><tr><th>Agent surface</th><th>Fixture</th></tr></thead><tbody><tr><td>Chat / domain assistant</td><td>A canonical prompt string, chosen once and frozen</td></tr><tr><td>Tool-heavy</td><td>Run <code>agentkit call &lt;tool&gt;</code> first to pin what the tool returns, then freeze the prompt that triggers it</td></tr><tr><td>GitHub webhook</td><td><code>agentkit github replay &lt;pr&gt; --events &#39;*&#39; --dry-run --out fixtures/github</code> snapshots real payloads for offline replay (<a href="./guides/github.html">GitHub guide</a>)</td></tr><tr><td>PR reviewer with host preparation</td><td>Diff, metadata, and gold labels pinned to commit SHAs; keep any live PR matrix small</td></tr><tr><td>Workspace-dependent</td><td><code>workspaceFiles</code> in <code>t.send</code> options, never developer-machine paths</td></tr></tbody></table><p>Tag the fast, reliably passing core <code>smoke</code> and run <code>--tag smoke</code> in the inner loop. Leave slow or flaky-prone cases untagged for explicit runs.</p><h3 id="materialize-api-backed-fixtures" tabindex="-1">Materialize API-backed fixtures <a class="header-anchor" href="#materialize-api-backed-fixtures" aria-label="Permalink to &quot;Materialize API-backed fixtures&quot;">​</a></h3><p>An input that only points at external data, such as a pull request URL, snapshot id, or pair of commit SHAs, is not self-contained. Fetch it once and commit the rendered fixture before you expand the suite.</p><ol><li>Save the diff, metadata, and labels under <code>fixtures/</code> at pinned revisions.</li><li>Seed those files with <code>workspaceFiles</code>, or read them from the fixture directory.</li><li>Assert decisions and output shape against the saved evidence.</li><li>Keep a small <code>smoke</code> subset for any remaining live pipeline checks.</li></ol><p><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 &quot;Keep improvements with regression evals&quot;">​</a></h2><p>Every <a href="./hillclimbing.html">hillclimb</a> round that keeps a change must land an eval that would have failed before the change. If you can&#39;t express the improvement as a gate (a <code>calledTool</code> shift, a bounded <code>action.result</code> count, an output-shape regex), the improvement is unverified, and it&#39;ll regress silently.</p><p>The rule cuts the other way too: never weaken an existing gate to make a round pass. That&#39;s the freeze line moving, and it turns your regression suite into a list of checks that no longer protect anything.</p><h2 id="compare-variants-on-live-traffic" tabindex="-1">Compare variants on live traffic <a class="header-anchor" href="#compare-variants-on-live-traffic" aria-label="Permalink to &quot;Compare variants on live traffic&quot;">​</a></h2><p>Use <code>defineAB</code> to compare variant metrics on live sessions. It is not a test runner and has no <code>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&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./ab.html">Live A/B metrics</a>: sticky variants and cumulative metrics on live sessions</li><li><a href="./hillclimbing.html">Hillclimbing</a>: the loop evals make trustworthy</li><li><a href="./building-with-agents.html">Building agents with agents</a>: have a coding agent write the first suite</li><li><a href="./guides/github.html">GitHub guide</a>: deterministic webhook fixtures with <code>github replay</code></li><li><a href="./reference/sessions.html">Sessions and streaming</a>: the events <code>t.events</code> contains</li></ul></div></div></main><footer class="VPDocFooter" data-v-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>
-    <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>
+<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 &quot;Run evals in the playground&quot;">​</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 (returns an <strong>Eval ID</strong> / <code>runId</code>), <code>GET /v1/dev/evals/runs/:runId</code> polls it, and <code>POST /v1/dev/evals/runs/:runId/cancel</code> cancels a running batch. 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>, <code>failed</code>, or <code>cancelled</code>. Configuration errors appear on a failed snapshot.</p><p>On <code>--prod</code> / <code>--url</code>, the CLI prints the Eval ID as soon as the batch is accepted (and a Playground deep link with <code>?view=evals&amp;evalRunId=…</code>):</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;"> eval</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --prod</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --slug</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> vulnerability-scanner</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --tag</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> deepsec</span></span>
+<span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># Eval ID: evalrun_…</span></span>
+<span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># Cancel:  agentkit eval cancel evalrun_… --prod --slug vulnerability-scanner</span></span>
+<span class="line"><span style="--shiki-light:#6A737D;--shiki-dark:#6A737D;"># Playground: https://…/playground?view=evals&amp;evalRunId=evalrun_…</span></span>
+<span class="line"></span>
+<span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> cancel</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> evalrun_…</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --prod</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --slug</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> vulnerability-scanner</span></span>
+<span class="line"><span style="--shiki-light:#6F42C1;--shiki-dark:#B392F0;">agentkit</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> eval</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> status</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> evalrun_…</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --prod</span><span style="--shiki-light:#005CC5;--shiki-dark:#79B8FF;"> --slug</span><span style="--shiki-light:#032F62;--shiki-dark:#9ECBFF;"> vulnerability-scanner</span></span></code></pre></div><p>The Evals tab prefers the server’s in-flight batch (<code>activeRunId</code>) over a stale tab-local remembered id, so CLI / Slack kicks show up without an incognito window.</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 &quot;What good cases assert&quot;">​</a></h2><p>Gate decisions and shape, not prose. Model wording varies run to run. Tool choice, tool avoidance, and output structure are the stable contract.</p><ol><li><code>t.succeeded()</code>: always, first.</li><li>The tool decision: <code>calledTool</code> for the intended path, <code>notCalledTool</code> for the likely wrong alternative. The pair is stronger than either alone.</li><li>Output shape: a regex for the contract (<code>/ready|blocked/i</code>, a JSON marker, a findings-block fence), never exact sentences.</li><li>For structured output, parse <code>t.reply</code> and check fields with <code>satisfies</code> instead of substring-matching JSON.</li></ol><p>The common failure modes: asserting exact phrasing, packing more than about five gates into one case (split it), and cases that depend on live external state that drifts (pin the input; see fixtures).</p><h2 id="pick-fixtures-by-agent-type" tabindex="-1">Pick fixtures by agent type <a class="header-anchor" href="#pick-fixtures-by-agent-type" aria-label="Permalink to &quot;Pick fixtures by agent type&quot;">​</a></h2><p>The right fixture depends on the surface under test.</p><table tabindex="0"><thead><tr><th>Agent surface</th><th>Fixture</th></tr></thead><tbody><tr><td>Chat / domain assistant</td><td>A canonical prompt string, chosen once and frozen</td></tr><tr><td>Tool-heavy</td><td>Run <code>agentkit call &lt;tool&gt;</code> first to pin what the tool returns, then freeze the prompt that triggers it</td></tr><tr><td>GitHub webhook</td><td><code>agentkit github replay &lt;pr&gt; --events &#39;*&#39; --dry-run --out fixtures/github</code> snapshots real payloads for offline replay (<a href="./guides/github.html">GitHub guide</a>)</td></tr><tr><td>PR reviewer with host preparation</td><td>Diff, metadata, and gold labels pinned to commit SHAs; keep any live PR matrix small</td></tr><tr><td>Workspace-dependent</td><td><code>workspaceFiles</code> in <code>t.send</code> options, never developer-machine paths</td></tr></tbody></table><p>Tag the fast, reliably passing core <code>smoke</code> and run <code>--tag smoke</code> in the inner loop. Leave slow or flaky-prone cases untagged for explicit runs.</p><h3 id="materialize-api-backed-fixtures" tabindex="-1">Materialize API-backed fixtures <a class="header-anchor" href="#materialize-api-backed-fixtures" aria-label="Permalink to &quot;Materialize API-backed fixtures&quot;">​</a></h3><p>An input that only points at external data, such as a pull request URL, snapshot id, or pair of commit SHAs, is not self-contained. Fetch it once and commit the rendered fixture before you expand the suite.</p><ol><li>Save the diff, metadata, and labels under <code>fixtures/</code> at pinned revisions.</li><li>Seed those files with <code>workspaceFiles</code>, or read them from the fixture directory.</li><li>Assert decisions and output shape against the saved evidence.</li><li>Keep a small <code>smoke</code> subset for any remaining live pipeline checks.</li></ol><p><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 &quot;Keep improvements with regression evals&quot;">​</a></h2><p>Every <a href="./hillclimbing.html">hillclimb</a> round that keeps a change must land an eval that would have failed before the change. If you can&#39;t express the improvement as a gate (a <code>calledTool</code> shift, a bounded <code>action.result</code> count, an output-shape regex), the improvement is unverified, and it&#39;ll regress silently.</p><p>The rule cuts the other way too: never weaken an existing gate to make a round pass. That&#39;s the freeze line moving, and it turns your regression suite into a list of checks that no longer protect anything.</p><h2 id="compare-variants-on-live-traffic" tabindex="-1">Compare variants on live traffic <a class="header-anchor" href="#compare-variants-on-live-traffic" aria-label="Permalink to &quot;Compare variants on live traffic&quot;">​</a></h2><p>Use <code>defineAB</code> to compare variant metrics on live sessions. It is not a test runner and has no <code>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&#39;s next <a class="header-anchor" href="#what-s-next" aria-label="Permalink to &quot;What&#39;s next&quot;">​</a></h2><p>Continue with these pages:</p><ul><li><a href="./ab.html">Live A/B metrics</a>: sticky variants and cumulative metrics on live sessions</li><li><a href="./hillclimbing.html">Hillclimbing</a>: the loop evals make trustworthy</li><li><a href="./building-with-agents.html">Building agents with agents</a>: have a coding agent write the first suite</li><li><a href="./guides/github.html">GitHub guide</a>: deterministic webhook fixtures with <code>github replay</code></li><li><a href="./reference/sessions.html">Sessions and streaming</a>: the events <code>t.events</code> contains</li></ul></div></div></main><footer class="VPDocFooter" data-v-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>
+    <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\":\"DYOjkRCX\",\"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\":\"DnYfr5V2\",\"reference_connections.md\":\"CsSr3rDz\",\"reference_hooks.md\":\"DyLVfE1O\",\"reference_http-api.md\":\"BncLd3PZ\",\"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>
     
   </body>
 </html>