premanmcp 1.0.5 → 1.1.0

package/bin/cli.js

@@ -36,6 +36,7 @@ import {
 import { HOSTED_HELP, linkCommand, runCommand, toolsCommand } from "./hosted.js";
 import { STATUS_HELP, statusCommand } from "./status.js";
 import { HOOK_HELP, hookCommand, installHook, scheduleHookRepair } from "./hook.js";
+import { EVAL_HELP } from "./eval.js";
 import { RUNNER_HELP, runnerCommand } from "./runner.js";
 import { VERIFY_HELP, verifyCommand } from "./verify.js";
 import {
@@ -94,6 +95,7 @@ function printHelp() {
     ["verify [options]", "Test endpoints against your local app"],
     ["hook install|uninstall|status", "Manage the git pre-push hook"],
     ["runner start|status|stop", "Run PreMan's queued agent work on this machine"],
+    ["eval run|doctor", "Run PreMan's queued agent eval against your local agent"],
     ["doctor", "Diagnose credentials, backend, target, integrations"],
     ["install-desktop", "Download and install the PreMan desktop app"],
     ["onboard", "Create an account, install the app signed in, then integrations"],
@@ -107,6 +109,7 @@ function printHelp() {
     ["", "Start the PreMan MCP server"],
     ["link|tools|run ...", "Drive a published hosted MCP"],
     ["endpoints list|discover|setup ...", "Discover, list, and set up API endpoints"],
+    ["test --agent [--only <id>]", "Measure your agent against the behaviours you turned on"],
     ["test <id> [--scenario ...] [--stress]", "Generate + run tests for an endpoint"],
     ["tests generate|review|setup|enrich", "Collections: generate, review, fixtures, enrich"],
   ];
@@ -119,7 +122,7 @@ function printHelp() {
 
 Usage:
 ${usageLines}
-${INTEGRATIONS_HELP}${CONNECT_HELP}${DISPATCH_HELP}${STATUS_HELP}${VERIFY_HELP}${HOOK_HELP}${RUNNER_HELP}${ACCOUNT_HELP}${DESKTOP_HELP}${ENDPOINTS_HELP}${TEST_HELP}${TESTS_HELP}
+${INTEGRATIONS_HELP}${CONNECT_HELP}${DISPATCH_HELP}${STATUS_HELP}${VERIFY_HELP}${HOOK_HELP}${RUNNER_HELP}${EVAL_HELP}${ACCOUNT_HELP}${DESKTOP_HELP}${ENDPOINTS_HELP}${TEST_HELP}${TESTS_HELP}
 Login options:
   --email <email>       Pre-fill the email prompt
   --backend <url>       PreMan backend URL. Defaults to ${DEFAULT_BACKEND}
@@ -259,6 +262,22 @@ async function main() {
     // depends on: the agent picker lives there, and importing it back would
     // close the cycle.
     await runnerCommand(commandArgs, { resolveAgent: resolveAgentForPairing });
+  } else if (command === "eval") {
+    // Same injection for the same reason, plus the runner protocol pieces:
+    // eval.js executes a leased run, and runner.js owns pairing and the stream.
+    // Passing them in keeps the dependency one-directional.
+    const { evalCommand } = await import("./eval.js");
+    const runner = await import("./runner.js");
+    await evalCommand(commandArgs, {
+      makeArgs,
+      registerRunner: runner.registerRunner,
+      readRunnerState: runner.readRunnerState,
+      saveRunnerState: runner.saveRunnerState,
+      deviceId: runner.deviceId,
+      runnerLoop: runner.runnerLoop,
+      pairingIsLive: runner.pairingIsLive,
+      resolveAgent: resolveAgentForPairing,
+    });
   } else if (command === "logout") {
     await logoutCommand();
   } else if (command === "doctor") {
@@ -297,6 +316,22 @@ async function main() {
     await runCommand(commandArgs);
   } else if (command === "endpoints") {
     await endpointsCommand(commandArgs);
+  } else if (command === "test" && commandArgs[0] === "--agent") {
+    // `test` already meant an endpoint, and it keeps meaning that. `--agent`
+    // names the other subject rather than inventing a second verb for it: the
+    // question is the same one, and which thing is under test is the part that
+    // differs. Routed here rather than inside testCommand because the two share
+    // no arguments beyond the word `test`, and matched in first position only so
+    // that an endpoint id is never overtaken by a flag further along the line.
+    const { agentTestCommand } = await import("./eval.js");
+    const runner = await import("./runner.js");
+    await agentTestCommand(commandArgs, {
+      makeArgs,
+      runnerLoop: runner.runnerLoop,
+      saveRunnerState: runner.saveRunnerState,
+      readRunnerState: runner.readRunnerState,
+      deviceId: runner.deviceId,
+    });
   } else if (command === "test") {
     await testCommand(commandArgs);
   } else if (command === "tests") {