@mindstudio-ai/agent 0.1.38 → 0.1.40

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/dist/cli.js CHANGED
@@ -4723,7 +4723,7 @@ function buildReferenceDocs(data) {
4723
4723
 
4724
4724
  Database queries (Query, Mutation) support both \`await\` and \`.then()/.catch()\` \u2014 use standard try/catch with await or chain \`.catch()\` directly.
4725
4725
 
4726
- Auth + DB identity: When an app has auth enabled, the authenticated user IS a row in the app's users table. \`auth.userId\` is the row's \`id\` \u2014 do NOT add a separate \`userId\` column. Access user data with \`Users.get(auth.userId)\`. The platform creates the user row on first login and manages the \`email\`, \`phone\`, and \`roles\` columns automatically.
4726
+ Auth + DB identity: When an app has auth enabled, the authenticated user IS a row in the app's users table. \`auth.userId\` is the row's \`id\` \u2014 do NOT add a separate \`userId\` column. Access user data with \`Users.get(auth.userId)\`. The platform creates the user row on first login and manages the \`email\`, \`phone\`, and \`roles\` columns automatically. IMPORTANT: The platform only populates the mapped auth columns (email, phone, roles) when creating the row \u2014 all other columns will be null until the developer's code sets them. Non-auth columns on the user table should be typed as optional (e.g. \`username?: string\`) and null-checked before use.
4727
4727
 
4728
4728
  Table options: \`db.defineTable<T>(name, { unique, defaults })\`.
4729
4729
  - \`unique: [['email'], ['userId', 'orgId']]\` \u2014 declares unique constraints (SDK communicates to platform, enables upsert).
@@ -4812,7 +4812,9 @@ var init_instructions = __esm({
4812
4812
  - **Never rank vendors** or claim one provider is better than another. Popularity reflects usage patterns, not quality judgments. Present options with their model IDs and let the caller choose. Never make assumptions about the capabilities of models, or suggest that a certain model is "good" at a specific task or better/worse than another at a task.
4813
4813
  - **Prefer popular models as defaults** in code examples unless the caller has a reason to use something specific.
4814
4814
  - **Always recommend latest-generation models**: For Anthropic, this is Claude 4 family, GPT-5 for OpenAI, Gemini 3 for Google, etc. MindStudio supports a ton of different models, including legacy models - but they are there for niche uses or backward compatibility/existing user requirements - they should NOT be used for new projects.
4815
- - Use the names of model input objects to infer capabilities. For example, any image generation model that supports source images in its config supports "remixing" or "image editing". Any video model with a start and end frame option supports creating "looping" videos. Do not look at model tags or descriptions to determine capabilities.
4815
+ - Use the names of model input objects to infer **technical capabilities** \u2014 e.g., a model with a source image input supports image editing, a video model with start/end frame supports looping. Do not look at model tags or descriptions to determine capabilities.
4816
+ - **Do NOT use config presets/enums to judge model quality or recommend models.** A model having style presets (e.g. "oil painting", "watercolor") does NOT make it better at those styles than models without presets. The recommended models (Seedream, Nano Banana Pro, etc.) are far more capable and achieve any style through prompting alone. Presets are a crutch for weaker models \u2014 never recommend a model just because it has an enum that matches the user's request.
4817
+ - **Never set maxResponseTokens to a low value.** Many models use extended thinking/reasoning that counts against the token limit. Setting maxResponseTokens to 256 or 512 will cause truncated or failed responses. Either omit it entirely (let the model's default handle it) or set it high (e.g. 16000+). Models stop generating on their own when the response is complete \u2014 a low cap does not save money, it just breaks output.
4816
4818
 
4817
4819
  ## Explicit preferences
4818
4820
  MindStudio has hundreds of models. Many of them are for niche use cases. In general, prefer to recommend the following models unless the user specifies otherwise.
@@ -5122,7 +5124,7 @@ async function startMcpServer(options) {
5122
5124
  capabilities: { tools: {} },
5123
5125
  serverInfo: {
5124
5126
  name: "mindstudio-agent",
5125
- version: "0.1.38"
5127
+ version: "0.1.40"
5126
5128
  },
5127
5129
  instructions: 'Welcome to MindStudio \u2014 a platform with 200+ AI models, 850+ third-party integrations, and pre-built agents.\n\nGetting started:\n1. Call `ask` with any question about the SDK \u2014 it knows every action, model, and connector and returns working code with real model IDs and config options. Examples: ask("generate an image with FLUX"), ask("what models support vision?"), ask("how do I send a Slack message?").\n2. Call `changeName` to set your display name \u2014 use your name or whatever your user calls you. This is how you\'ll appear in MindStudio request logs.\n3. If you have a profile picture or icon, call `uploadFile` to upload it, then `changeProfilePicture` with the returned URL.\n4. For manual browsing, call `listActions` to discover all available actions.\n\nThen use the tools to generate text, images, video, audio, search the web, work with data sources, run agents, and more.\n\nImportant:\n- AI-powered actions (text generation, image generation, video, audio, etc.) cost money. Before running these, call `estimateActionCost` and confirm with the user before proceeding \u2014 unless they\'ve explicitly told you to go ahead.\n- Not all agents from `listAgents` are configured for API use. Do not try to run an agent just because it appears in the list \u2014 it will likely fail. Only run agents the user specifically asks you to run.'
5128
5130
  });
@@ -6079,7 +6081,7 @@ function isNewerVersion(current, latest) {
6079
6081
  return false;
6080
6082
  }
6081
6083
  async function checkForUpdate() {
6082
- const currentVersion = "0.1.38";
6084
+ const currentVersion = "0.1.40";
6083
6085
  if (!currentVersion) return null;
6084
6086
  try {
6085
6087
  const { loadConfig: loadConfig2, saveConfig: saveConfig2 } = await Promise.resolve().then(() => (init_config(), config_exports));
@@ -6108,7 +6110,7 @@ async function checkForUpdate() {
6108
6110
  }
6109
6111
  }
6110
6112
  function printUpdateNotice(latestVersion) {
6111
- const currentVersion = "0.1.38";
6113
+ const currentVersion = "0.1.40";
6112
6114
  process.stderr.write(
6113
6115
  `
6114
6116
  ${ansi2.cyanBright("Update available")} ${ansi2.gray(currentVersion + " \u2192")} ${ansi2.cyanBold(latestVersion)}
@@ -6121,7 +6123,7 @@ function isStandaloneBinary() {
6121
6123
  return !argv1.includes("node_modules");
6122
6124
  }
6123
6125
  async function cmdUpdate() {
6124
- const currentVersion = "0.1.38";
6126
+ const currentVersion = "0.1.40";
6125
6127
  process.stderr.write(
6126
6128
  ` ${ansi2.gray("Current version:")} ${currentVersion}
6127
6129
  `
@@ -6256,7 +6258,7 @@ async function cmdLogin(options) {
6256
6258
  process.stderr.write("\n");
6257
6259
  printLogo();
6258
6260
  process.stderr.write("\n");
6259
- const ver = "0.1.38";
6261
+ const ver = "0.1.40";
6260
6262
  process.stderr.write(
6261
6263
  ` ${ansi2.bold("MindStudio Agent")} ${ver ? " " + ansi2.gray("v" + ver) : ""}
6262
6264
  `
@@ -6583,7 +6585,7 @@ async function main() {
6583
6585
  try {
6584
6586
  if (command === "version" || command === "-v") {
6585
6587
  process.stdout.write(
6586
- "0.1.38\n"
6588
+ "0.1.40\n"
6587
6589
  );
6588
6590
  return;
6589
6591
  }