ur-agent 1.37.1 → 1.37.3

package/docs/IDE.md

@@ -86,9 +86,14 @@ Install it with `ur ide install` (offers the bundled VSIX) or from the
 packaged `.vsix`. The extension is bundled inside this repository and
 packaged as a local VSIX when installed from UR-AGENT; the public install
 path does not depend on an unpublished marketplace extension ID. Every
-feature below requires the UR CLI on your `PATH` — the extension calls `ur`
-as a local subprocess and never talks to a model provider or network service
-directly.
+feature below runs UR as a local subprocess and never talks to a model
+provider or network service directly.
+
+When the workspace itself is a UR checkout, the extension prefers the
+workspace-local runtime (`bun dist/cli.js`, then `node bin/ur.js`) before
+falling back to `ur` from `PATH`. Set `ur.executablePath` in VS Code settings
+to force a specific binary; `ur.executableArgs` can provide fixed wrapper
+arguments before normal UR command arguments.
 
 ### Chat panel
 

package/docs/providers.md

@@ -420,3 +420,32 @@ Local/server providers use their normal endpoints:
 - LM Studio: `http://localhost:1234/v1`
 - llama.cpp server mode: `http://localhost:8080/v1`
 - vLLM server mode: `http://localhost:8000/v1`
+
+## Optional Live Provider Smoke
+
+`bun run provider:smoke` runs optional live checks only for providers with the
+required environment variables present. With no variables set it exits
+successfully and reports every provider as skipped, so CI does not need secrets.
+
+Configured providers run one short text request, a finite-timeout streaming
+request, and, when `PROVIDER_SMOKE_TOOL_CALLS=1`, a forced tool-call request.
+
+Required variables:
+
+| Provider | Required env vars | Optional env vars |
+| --- | --- | --- |
+| OpenAI-compatible | `OPENAI_COMPATIBLE_BASE_URL`, `OPENAI_COMPATIBLE_MODEL` | `OPENAI_COMPATIBLE_API_KEY` |
+| OpenRouter | `OPENROUTER_API_KEY`, `OPENROUTER_MODEL` |  |
+| Anthropic | `ANTHROPIC_API_KEY`, `ANTHROPIC_MODEL` | `ANTHROPIC_BASE_URL` |
+| Gemini | `GEMINI_API_KEY`, `GEMINI_MODEL` | `GEMINI_BASE_URL` |
+| Ollama | `OLLAMA_MODEL` | `OLLAMA_BASE_URL` or `OLLAMA_HOST` |
+| LM Studio | `LMSTUDIO_BASE_URL`, `LMSTUDIO_MODEL` | `LMSTUDIO_API_KEY` |
+| vLLM | `VLLM_BASE_URL`, `VLLM_MODEL` | `VLLM_API_KEY` |
+
+Common knobs:
+
+```sh
+PROVIDER_SMOKE_TIMEOUT_MS=30000
+PROVIDER_SMOKE_MAX_RETRIES=0
+PROVIDER_SMOKE_TOOL_CALLS=1
+```

package/documentation/index.html

@@ -44,7 +44,7 @@
     <main id="content" class="content">
       <header class="topbar">
         <div>
-          <p class="eyebrow">Version 1.37.0</p>
+          <p class="eyebrow">Version 1.37.3</p>
           <h1>UR-AGENT Documentation</h1>
           <p class="lead">A practical, tutorial-style reference for installing, configuring, automating, extending, and operating UR-AGENT.</p>
         </div>
@@ -58,12 +58,12 @@
         <div class="section-heading">
           <p class="eyebrow">What UR Is</p>
           <h2>Autonomous engineering workflow engine</h2>
-          <p>UR is not only chat, autocomplete, or code edits. It turns work into a plan, executes patches, tests and verifies them, documents the result, benchmarks behavior, and keeps the run reproducible from the terminal.</p>
+          <p>UR is not only chat, autocomplete, or code edits. It turns work into a plan, executes patches, tests and verifies them, documents the result, benchmarks behavior, and keeps the run reproducible.</p>
         </div>
         <div class="stat-grid">
           <article>
             <strong>Local-first runtime</strong>
-            <span>Use local Ollama/OpenAI-compatible runtimes for private and offline work, or explicitly select configured API provider paths. Subscription access stays visible but unavailable when no independent backend exists.</span>
+            <span>Use local Ollama/OpenAI-compatible runtimes for private and offline work, or explicitly select configured API provider paths. Subscription access stays visible but unavailable when not authenticated.</span>
           </article>
           <article>
             <strong>Interactive and headless</strong>
@@ -128,7 +128,7 @@ ur -p --output-format json "Review the current diff"</code></pre>
         </div>
         <div class="callout">
           <strong>Runtime requirement</strong>
-          <p>UR expects Bun. For local-first work, run a local Ollama app/server at <code>http://localhost:11434/api</code>. Set <code>OLLAMA_MODEL</code> or <code>UR_MODEL</code> to force a model, or use <code>ur provider</code> and <code>ur config set</code> to select official subscription, API, or OpenAI-compatible providers.</p>
+          <p>UR expects Bun. For local-first work, run a local Ollama app/server at <code>http://localhost:11434/api</code>. Set <code>OLLAMA_MODEL</code> or <code>UR_MODEL</code> to force a model route.</p>
         </div>
       </section>
 
@@ -136,7 +136,7 @@ ur -p --output-format json "Review the current diff"</code></pre>
         <div class="section-heading">
           <p class="eyebrow">Legal Access Paths</p>
           <h2>Providers and authentication</h2>
-          <p>UR-AGENT integrates official access paths only. It never scrapes browser sessions, extracts OAuth tokens, reads hidden provider auth files, bypasses provider restrictions, or proxies consumer web sessions as APIs.</p>
+          <p>UR-AGENT integrates official access paths only. It never scrapes browser sessions, extracts OAuth tokens, reads hidden provider auth files, bypasses provider restrictions, or proxies credentials.</p>
         </div>
         <div class="reference-grid">
           <article>
@@ -146,7 +146,7 @@ ur auth claude
 ur auth gemini
 ur auth antigravity
 ur provider doctor agy</code></pre>
-            <p>Codex CLI, Claude Code, Gemini CLI, and Antigravity are first-class providers that dispatch each turn through the vendor's official CLI using your subscription login. They are optional, never required, and never used as a silent fallback. The generic subscription entry is an internal placeholder hidden from listings and exposes no fake models.</p>
+            <p>Codex CLI, Claude Code, Gemini CLI, and Antigravity are first-class providers that dispatch each turn through the vendor's official CLI using your subscription login. They are optional and can be disabled.</p>
           </article>
           <article>
             <h3>API and local providers</h3>
@@ -155,12 +155,12 @@ ur config set provider openai-api
 ur config set base_url http://localhost:11434/v1
 ur config set model qwen3-coder:480b-cloud
 ur config set provider.fallback ollama</code></pre>
-            <p>API providers require explicit selection and read keys from a key stored via <code>ur connect</code> (OS keychain) or from environment variables. Local providers include Ollama, LM Studio, llama.cpp server mode, and vLLM OpenAI-compatible server mode. Subscription CLI providers (Codex CLI, Claude Code, Gemini CLI, Antigravity) dispatch through the official vendor CLIs.</p>
+            <p>API providers require explicit selection and read keys from a key stored via <code>ur connect</code> (OS keychain) or from environment variables. Local providers include Ollama, LM Studio, llama.cpp, and vLLM.</p>
           </article>
           <article>
             <h3>Status bar and updates</h3>
-            <pre><code>Ollama | llama3 | ask | main | update 1.37.0 available</code></pre>
-            <p>The interactive status bar shows only important runtime state: provider, model, mode, branch, active tasks, checks status when known, and update availability. It is hidden in CI, dumb terminals, and non-interactive mode.</p>
+            <pre><code>Ollama | llama3 | ask | main | update 1.37.3 available</code></pre>
+            <p>The interactive status bar shows only important runtime state: provider, model, mode, branch, active tasks, checks status when known, and update availability. It is hidden in CI, dumb terminals, and print mode.</p>
           </article>
         </div>
       </section>
@@ -169,7 +169,7 @@ ur config set provider.fallback ollama</code></pre>
         <div class="section-heading">
           <p class="eyebrow">Runtime</p>
           <h2>Architecture</h2>
-          <p>The CLI is the control plane. Project context and <code>.ur</code> assets feed agent orchestration, which uses approved provider adapters, local-first model runtimes, and approved tools to produce code, docs, PRs, reports, and automation outputs.</p>
+          <p>The CLI is the control plane. Project context and <code>.ur</code> assets feed agent orchestration, which uses approved provider adapters, local-first model runtimes, and approved tools to produce auditable results.</p>
         </div>
         <figure class="diagram">
           <img src="assets/ur-architecture.svg" alt="UR-AGENT runtime architecture diagram">

package/extensions/vscode-ur-inline-diffs/README.md

@@ -11,7 +11,13 @@ marketplace extension ID.
 
 The extension never talks to a model provider or network service directly —
 every AI request goes through your local `ur` CLI, the same way `ur` behaves
-in a terminal. Everything below requires the UR CLI on your `PATH`.
+in a terminal.
+
+When the opened workspace is a UR checkout, the extension prefers the
+workspace-local runtime (`bun dist/cli.js`, then `node bin/ur.js`) before
+falling back to `ur` on `PATH`. To force a specific binary, set
+`ur.executablePath` in VS Code settings. Use `ur.executableArgs` only for
+custom wrappers that need fixed arguments before normal UR arguments.
 
 ## Chat
 

package/extensions/vscode-ur-inline-diffs/out/extension.js

@@ -221,23 +221,66 @@ var DiffTreeProvider = class {
 };
 
 // src/bridge/urCli.ts
-var import_node_child_process = require("node:child_process");
+var import_node_child_process2 = require("node:child_process");
 var import_node_util = require("node:util");
-var execFileAsync = (0, import_node_util.promisify)(import_node_child_process.execFile);
+
+// src/bridge/urCommand.ts
+var import_node_child_process = require("node:child_process");
+var import_node_fs = require("node:fs");
+var import_node_path = require("node:path");
+function resolveUrCommand(options) {
+  const config = options.config ?? {};
+  const configuredPath = config.executablePath?.trim();
+  const configuredArgs = normalizeExecutableArgs(config.executableArgs);
+  if (configuredPath) {
+    return command(configuredPath, configuredArgs, "configured");
+  }
+  const pathExists = options.pathExists ?? import_node_fs.existsSync;
+  const bunCommand = options.bunCommand ?? "bun";
+  const nodeCommand = options.nodeCommand ?? "node";
+  const distEntrypoint = (0, import_node_path.join)(options.cwd, "dist", "cli.js");
+  if (pathExists(distEntrypoint) && (options.bunAvailable ?? isBunAvailable)(bunCommand)) {
+    return command(bunCommand, [distEntrypoint], "workspace-dist");
+  }
+  const launcherEntrypoint = (0, import_node_path.join)(options.cwd, "bin", "ur.js");
+  if (pathExists(launcherEntrypoint)) {
+    return command(nodeCommand, [launcherEntrypoint], "workspace-launcher");
+  }
+  return command("ur", [], "path");
+}
+function formatResolvedUrCommand(resolved) {
+  return [resolved.command, ...resolved.args].join(" ");
+}
+function command(commandName, args, source) {
+  const resolved = { command: commandName, args, source, display: "" };
+  return { ...resolved, display: formatResolvedUrCommand(resolved) };
+}
+function normalizeExecutableArgs(args) {
+  return Array.isArray(args) ? args.filter((arg) => typeof arg === "string" && arg.length > 0) : [];
+}
+function isBunAvailable(commandName) {
+  const result = (0, import_node_child_process.spawnSync)(commandName, ["--version"], { stdio: "ignore" });
+  return !result.error && result.status === 0;
+}
+
+// src/bridge/urCli.ts
+var execFileAsync = (0, import_node_util.promisify)(import_node_child_process2.execFile);
 async function runUrCli(args, options) {
+  const executable = resolveUrCommand({ cwd: options.cwd, config: readUrCommandConfig() });
   try {
-    const { stdout, stderr } = await execFileAsync("ur", args, {
+    const { stdout, stderr } = await execFileAsync(executable.command, [...executable.args, ...args], {
       cwd: options.cwd,
       shell: false
     });
     return { stdout, stderr };
   } catch (error) {
-    throw new Error(formatUrCliError(args, error));
+    throw new Error(formatUrCliError(args, error, executable, options.cwd));
   }
 }
 async function runUrCliCapture(args, options) {
+  const executable = resolveUrCommand({ cwd: options.cwd, config: readUrCommandConfig() });
   try {
-    const { stdout, stderr } = await execFileAsync("ur", args, {
+    const { stdout, stderr } = await execFileAsync(executable.command, [...executable.args, ...args], {
       cwd: options.cwd,
       shell: false
     });
@@ -246,13 +289,34 @@ async function runUrCliCapture(args, options) {
     if (isCapturedNonZeroExit(error)) {
       return { stdout: error.stdout, stderr: error.stderr, exitCode: error.code };
     }
-    throw new Error(formatUrCliError(args, error));
+    throw new Error(formatUrCliError(args, error, executable, options.cwd));
   }
 }
-function formatUrCliError(args, error) {
+function readUrCommandConfig() {
+  try {
+    const vscode18 = require("vscode");
+    const config = vscode18.workspace.getConfiguration("ur");
+    const executablePath = config.get("executablePath")?.trim();
+    const executableArgs = config.get("executableArgs") ?? [];
+    return {
+      executablePath: executablePath || void 0,
+      executableArgs: executableArgs.filter((arg) => typeof arg === "string" && arg.length > 0)
+    };
+  } catch {
+    return {};
+  }
+}
+function formatUrCliError(args, error, executable, cwd) {
   const stderr = hasStderr(error) ? error.stderr.trim() : "";
   const detail = stderr || (error instanceof Error ? error.message : String(error));
-  return `Failed to run \`ur ${args.join(" ")}\`: ${detail}. Ensure the UR CLI is installed and on PATH.`;
+  return [
+    `Failed to run UR command.`,
+    `Executable: ${executable.display}`,
+    `cwd: ${cwd}`,
+    `Args: ${args.join(" ")}`,
+    `Error: ${detail}`,
+    `Hint: Set ur.executablePath in VS Code settings if the extension is using the wrong UR binary.`
+  ].join("\n");
 }
 function hasStderr(error) {
   return typeof error === "object" && error !== null && "stderr" in error && typeof error.stderr === "string";
@@ -409,7 +473,7 @@ function isCanUseToolRequest(message) {
 }
 
 // src/bridge/urProcess.ts
-var import_node_child_process2 = require("node:child_process");
+var import_node_child_process3 = require("node:child_process");
 var NdjsonBuffer = class {
   buffer = "";
   /** Feed a raw chunk (may contain zero, one, or many complete lines, and may
@@ -477,14 +541,14 @@ function buildControlResponse(requestId, decision) {
     }
   };
 }
-var defaultSpawn = (command, args, options) => (0, import_node_child_process2.spawn)(command, args, options);
+var defaultSpawn = (command2, args, options) => (0, import_node_child_process3.spawn)(command2, args, options);
 function runUrTurn(request, handlers, deps = {}) {
   const spawnFn = deps.spawn ?? defaultSpawn;
-  const command = deps.command ?? "ur";
-  const args = buildUrArgs(request);
+  const executable = deps.executable ?? (deps.command ? { command: deps.command, args: [], source: "configured", display: deps.command } : resolveUrCommand({ cwd: request.cwd }));
+  const args = [...executable.args, ...buildUrArgs(request)];
   let child;
   try {
-    child = spawnFn(command, args, {
+    child = spawnFn(executable.command, args, {
       cwd: request.cwd,
       shell: false,
       stdio: ["pipe", "pipe", "pipe"]
@@ -493,10 +557,18 @@ function runUrTurn(request, handlers, deps = {}) {
     handlers.onExit({
       ok: false,
       exitCode: null,
+      signal: null,
       canceled: false,
       sawResult: false,
       stderr: "",
-      error: `Failed to start \`${command}\`: ${errorMessage2(error)}. Ensure the UR CLI is installed and on PATH.`
+      error: formatTurnFailure({
+        executable,
+        cwd: request.cwd,
+        exitCode: null,
+        signal: null,
+        stderr: "",
+        reason: `Failed to start: ${errorMessage2(error)}`
+      })
     });
     return { cancel: () => {
     } };
@@ -507,7 +579,7 @@ function runUrTurn(request, handlers, deps = {}) {
   let resultIsError = false;
   let canceled = false;
   let settled = false;
-  const finish = (exitCode, spawnError) => {
+  const finish = (exitCode, signal, spawnError) => {
     if (settled) return;
     settled = true;
     const stderr = stderrChunks.join("");
@@ -515,10 +587,11 @@ function runUrTurn(request, handlers, deps = {}) {
     handlers.onExit({
       ok,
       exitCode,
+      signal,
       canceled,
       sawResult,
       stderr,
-      error: spawnError ?? (!ok && !canceled ? deriveErrorMessage(sawResult, resultIsError, exitCode, stderr) : void 0)
+      error: spawnError ?? (!ok && !canceled ? deriveErrorMessage(executable, request.cwd, sawResult, resultIsError, exitCode, signal, stderr) : void 0)
     });
   };
   const handleMessage = (message) => {
@@ -547,13 +620,24 @@ function runUrTurn(request, handlers, deps = {}) {
     stderrChunks.push(chunk.toString("utf8"));
   });
   child.on("error", (error) => {
-    finish(null, `Failed to run \`${command}\`: ${errorMessage2(error)}. Ensure the UR CLI is installed and on PATH.`);
+    finish(
+      null,
+      null,
+      formatTurnFailure({
+        executable,
+        cwd: request.cwd,
+        exitCode: null,
+        signal: null,
+        stderr: stderrChunks.join(""),
+        reason: `Failed to run: ${errorMessage2(error)}`
+      })
+    );
   });
-  child.on("exit", (code) => {
+  child.on("exit", (code, signal) => {
     for (const message of stdoutBuffer.flush()) {
       handleMessage(message);
     }
-    finish(code);
+    finish(code, signal);
   });
   return {
     cancel: () => {
@@ -570,11 +654,26 @@ function writeControlResponse(child, requestId, decision) {
   } catch {
   }
 }
-function deriveErrorMessage(sawResult, resultIsError, exitCode, stderr) {
-  if (sawResult && resultIsError) return "UR reported an error completing this turn.";
-  const trimmedStderr = stderr.trim();
-  if (trimmedStderr) return trimmedStderr;
-  return `UR exited with code ${exitCode ?? "unknown"} and produced no result.`;
+function deriveErrorMessage(executable, cwd, sawResult, resultIsError, exitCode, signal, stderr) {
+  const reason = sawResult && resultIsError ? "UR reported an error completing this turn." : "UR exited without producing a successful result.";
+  return formatTurnFailure({ executable, cwd, exitCode, signal, stderr, reason });
+}
+function formatTurnFailure(options) {
+  const stderr = summarizeStderr(options.stderr);
+  return [
+    `UR chat backend failed.`,
+    `Executable: ${options.executable.display}`,
+    `cwd: ${options.cwd}`,
+    `Exit: code ${options.exitCode ?? "unknown"}, signal ${options.signal ?? "none"}`,
+    `stderr: ${stderr || "<empty>"}`,
+    options.reason,
+    `Hint: Set ur.executablePath in VS Code settings if the extension is using the wrong UR binary.`
+  ].join("\n");
+}
+function summarizeStderr(stderr) {
+  const trimmed = stderr.trim();
+  if (trimmed.length <= 2e3) return trimmed;
+  return `${trimmed.slice(0, 2e3)}\u2026`;
 }
 function errorMessage2(error) {
   return error instanceof Error ? error.message : String(error);
@@ -1367,7 +1466,8 @@ var ChatController = class {
         onMessage: (message) => this.handleStreamMessage(root, sessionId, message),
         onControlRequest: (request) => this.handlePermissionRequest(request),
         onExit: (result) => this.handleTurnExit(root, result)
-      }
+      },
+      { executable: resolveUrCommand({ cwd: root, config: readUrCommandConfig() }) }
     );
   }
   handleStreamMessage(root, sessionId, message) {
@@ -1489,13 +1589,13 @@ var ChatController = class {
 // src/chat/chatTreeProvider.ts
 var vscode7 = __toESM(require("vscode"));
 var ActionItem = class extends vscode7.TreeItem {
-  constructor(label, description, icon, command, tooltip) {
+  constructor(label, description, icon, command2, tooltip) {
     super(label, vscode7.TreeItemCollapsibleState.None);
     this.contextValue = "urChatAction";
     this.description = description;
     this.iconPath = new vscode7.ThemeIcon(icon);
     this.tooltip = tooltip ?? `${label}${description ? ` - ${description}` : ""}`;
-    this.command = command;
+    this.command = command2;
   }
 };
 var InfoItem3 = class extends vscode7.TreeItem {
@@ -1567,11 +1667,11 @@ var ChatTreeProvider = class {
 };
 
 // src/diffs/actions.ts
-var import_node_child_process3 = require("node:child_process");
+var import_node_child_process4 = require("node:child_process");
 var fs3 = __toESM(require("node:fs"));
 var import_node_util2 = require("node:util");
 var vscode8 = __toESM(require("vscode"));
-var execFileAsync2 = (0, import_node_util2.promisify)(import_node_child_process3.execFile);
+var execFileAsync2 = (0, import_node_util2.promisify)(import_node_child_process4.execFile);
 async function commentDiff(item, provider) {
   const root = workspaceRoot();
   const bundle = item?.bundle;
@@ -2181,7 +2281,7 @@ async function showAgentOptions(cwd) {
 }
 
 // src/review/reviewDiff.ts
-var import_node_child_process4 = require("node:child_process");
+var import_node_child_process5 = require("node:child_process");
 var import_node_util3 = require("node:util");
 var vscode13 = __toESM(require("vscode"));
 
@@ -2199,7 +2299,7 @@ function buildReviewPrompt(diff) {
 }
 
 // src/review/reviewDiff.ts
-var execFileAsync3 = (0, import_node_util3.promisify)(import_node_child_process4.execFile);
+var execFileAsync3 = (0, import_node_util3.promisify)(import_node_child_process5.execFile);
 async function captureGitDiff(cwd) {
   const { stdout } = await execFileAsync3("git", ["diff", "HEAD"], { cwd, shell: false });
   return stdout;

package/extensions/vscode-ur-inline-diffs/package.json

@@ -2,7 +2,7 @@
   "name": "ur-inline-diffs",
   "displayName": "UR Inline Diffs",
   "description": "Review, apply, and reject UR inline diff bundles from .ur/ide/diffs inside VS Code.",
-  "version": "1.37.1",
+  "version": "1.37.3",
   "publisher": "ur-agent",
   "engines": {
     "vscode": "^1.92.0"
@@ -55,6 +55,24 @@
     "typescript": "^5.7.0"
   },
   "contributes": {
+    "configuration": {
+      "title": "UR",
+      "properties": {
+        "ur.executablePath": {
+          "type": "string",
+          "default": "",
+          "description": "Absolute path or command name for the UR executable the extension should run. Leave empty to prefer a repo-local UR checkout in the workspace, then fall back to ur on PATH."
+        },
+        "ur.executableArgs": {
+          "type": "array",
+          "default": [],
+          "items": {
+            "type": "string"
+          },
+          "description": "Optional arguments inserted after ur.executablePath and before UR command arguments. Use this only for custom wrappers, for example when ur.executablePath points to bun and the first arg is a dist/cli.js path."
+        }
+      }
+    },
     "viewsContainers": {
       "activitybar": [
         {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ur-agent",
-  "version": "1.37.1",
+  "version": "1.37.3",
   "description": "UR-AGENT \u2014 autonomous engineering workflow engine (plan, execute, test, verify, document, benchmark, reproduce)",
   "type": "module",
   "packageManager": "bun@1.3.14",
@@ -49,11 +49,23 @@
     "bundle": "node scripts/bundle.mjs",
     "build": "node scripts/bundle.mjs",
     "release:check": "node scripts/release-check.mjs",
-    "typecheck": "bun run tsc --noEmit",
+    "release:create-source-zip": "bun scripts/source-zip-create.mjs",
+    "release:check-source-zip": "node scripts/source-zip-check.mjs",
+    "typecheck": "bun run tsc --noEmit && bun run typecheck:strict-core",
+    "typecheck:strict-core": "node scripts/strict-core-check.mjs",
     "lint": "node scripts/lint.mjs",
     "test": "bun test",
     "smoke": "node ./bin/ur.js --version",
     "secrets:scan": "bash scripts/secret-scan.sh",
+    "benchmark:report": "node scripts/benchmark-report.mjs",
+    "benchmark:smoke": "node scripts/benchmark-smoke.mjs",
+    "benchmark:local": "node scripts/benchmark-local.mjs",
+    "benchmark:compare": "node scripts/benchmark-compare.mjs",
+    "benchmark:swe-bench-lite": "node scripts/benchmark-external.mjs swe-bench-lite",
+    "benchmark:terminal-bench": "node scripts/benchmark-external.mjs terminal-bench",
+    "benchmark:aider-polyglot": "node scripts/benchmark-external.mjs aider-polyglot",
+    "safety:matrix": "bun scripts/generate-safety-matrix.mjs",
+    "provider:smoke": "bun scripts/provider-smoke.mjs",
     "package:check": "node scripts/package-check.mjs",
     "prepack": "bun run release:check"
   },