@agentmark-ai/cli 0.10.0 → 0.10.1

package/README.md

@@ -1 +1,143 @@
-# Agentmark CLI
+# AgentMark CLI
+
+The command-line tool for developing, testing, and evaluating AI agents with AgentMark.
+
+## Installation
+
+```bash
+npm install -g @agentmark-ai/cli
+```
+
+Or use it directly with `npx`:
+
+```bash
+npx @agentmark-ai/cli dev
+```
+
+## Quick Start
+
+```bash
+# Scaffold a new project
+npm create agentmark@latest
+
+# Start the dev server (API + trace UI + hot reload)
+agentmark dev
+
+# Run a prompt with its test props
+agentmark run-prompt my-prompt.prompt.mdx
+
+# Run an experiment against a dataset
+agentmark run-experiment my-prompt.prompt.mdx
+```
+
+## Commands
+
+### `agentmark dev`
+
+Start the local development environment: API server, webhook server, and trace UI.
+
+```bash
+agentmark dev
+agentmark dev --api-port 9418 --app-port 3000
+agentmark dev --remote    # Connect to AgentMark Cloud (login + trace forwarding)
+agentmark dev --tunnel    # Expose webhook server publicly
+```
+
+The dev server auto-detects your project language (TypeScript or Python), finds your virtual environment, and handles port conflicts automatically.
+
+### `agentmark run-prompt <filepath>`
+
+Execute a single prompt and display the result.
+
+```bash
+# Run with test props from the prompt's frontmatter
+agentmark run-prompt customer-support.prompt.mdx
+
+# Run with custom props
+agentmark run-prompt customer-support.prompt.mdx --props '{"customer_question": "Where is my order?"}'
+
+# Run with props from a file
+agentmark run-prompt customer-support.prompt.mdx --props-file test-data.json
+```
+
+Output includes the LLM response, token usage, cost, and a link to the trace in the local UI.
+
+### `agentmark run-experiment <filepath>`
+
+Run a prompt against every item in its dataset, with optional evaluations.
+
+```bash
+# Run with evals (default)
+agentmark run-experiment my-prompt.prompt.mdx
+
+# Skip evals
+agentmark run-experiment my-prompt.prompt.mdx --skip-eval
+
+# Output as JSON instead of table
+agentmark run-experiment my-prompt.prompt.mdx --format json
+
+# Fail if pass rate is below 80%
+agentmark run-experiment my-prompt.prompt.mdx --threshold 80
+```
+
+Output formats: `table` (default), `csv`, `json`, `jsonl`.
+
+### `agentmark build`
+
+Pre-compile `.prompt.mdx` files into JSON for production use with the file loader.
+
+```bash
+agentmark build
+agentmark build --out dist/prompts
+```
+
+### `agentmark generate-types`
+
+Generate TypeScript type definitions from your prompts for type-safe usage in code.
+
+```bash
+agentmark generate-types
+```
+
+### `agentmark generate-schema`
+
+Generate a JSON Schema for `.prompt.mdx` frontmatter, enabling IDE validation and autocomplete.
+
+```bash
+agentmark generate-schema
+agentmark generate-schema --out .agentmark
+```
+
+### `agentmark pull-models`
+
+Interactively select and add LLM models from a provider to your `agentmark.json`.
+
+```bash
+agentmark pull-models
+```
+
+### `agentmark login` / `agentmark logout`
+
+Authenticate with AgentMark Cloud.
+
+```bash
+agentmark login
+agentmark logout
+```
+
+### `agentmark link`
+
+Link your project to an AgentMark Cloud app for trace forwarding.
+
+```bash
+agentmark link
+agentmark link --app-id <uuid>
+```
+
+## Documentation
+
+Full documentation at [docs.agentmark.co](https://docs.agentmark.co/agentmark/).
+
+## License
+
+[MIT](../../LICENSE.md)

package/dist/.next/BUILD_ID

@@ -1 +1 @@
-p7ZqHwMUDueov4RV9Be-X
+2NZm040KzCzDPOX6MqoKG

package/dist/.next/app-build-manifest.json

@@ -75,7 +75,7 @@
       "static/chunks/352-d7aa9d2da963e9e5.js",
       "static/chunks/49-6efbcf2b4f20d950.js",
       "static/chunks/349-9266e48c00c8ac76.js",
-      "static/chunks/41-8ba62a009d80b6dc.js",
+      "static/chunks/41-f512c46d7ea943f6.js",
       "static/chunks/app/requests/page-aa9725ed8b67059b.js"
     ],
     "/sessions/page": [
@@ -92,7 +92,7 @@
       "static/chunks/352-d7aa9d2da963e9e5.js",
       "static/chunks/49-6efbcf2b4f20d950.js",
       "static/chunks/349-9266e48c00c8ac76.js",
-      "static/chunks/41-8ba62a009d80b6dc.js",
+      "static/chunks/41-f512c46d7ea943f6.js",
       "static/chunks/app/sessions/page-bf2019c391763736.js"
     ],
     "/traces/page": [
@@ -109,7 +109,7 @@
       "static/chunks/352-d7aa9d2da963e9e5.js",
       "static/chunks/49-6efbcf2b4f20d950.js",
       "static/chunks/349-9266e48c00c8ac76.js",
-      "static/chunks/41-8ba62a009d80b6dc.js",
+      "static/chunks/41-f512c46d7ea943f6.js",
       "static/chunks/app/traces/page-c43b6aa5cd13fdf1.js"
     ]
   }

package/dist/.next/build-manifest.json

@@ -5,8 +5,8 @@
   "devFiles": [],
   "ampDevFiles": [],
   "lowPriorityFiles": [
-    "static/p7ZqHwMUDueov4RV9Be-X/_buildManifest.js",
-    "static/p7ZqHwMUDueov4RV9Be-X/_ssgManifest.js"
+    "static/2NZm040KzCzDPOX6MqoKG/_buildManifest.js",
+    "static/2NZm040KzCzDPOX6MqoKG/_ssgManifest.js"
   ],
   "rootMainFiles": [
     "static/chunks/webpack-f10b0ea54196b48e.js",

package/dist/.next/cache/.previewinfo

@@ -1 +1 @@
-{"previewModeId":"6fd364d8232f1bf4865e801056a55d10","previewModeSigningKey":"287680828a93619def2a62f48485360b6203ae496e39d25060b410106ba38020","previewModeEncryptionKey":"d88662eb87cae7c69d67d03b707252641137370b45fb04b1843c8c600f4fec43","expireAt":1773770547134}
+{"previewModeId":"3d2ff63eaf01f7abb0ba0caeb3c17d0c","previewModeSigningKey":"981a172e225e7b17909282f71dbf463040b54bf9bf7eba4644fd99aa08e76454","previewModeEncryptionKey":"aca121c70f6014285fd5e771c8346ae5ee08d3e6a8a14de58b70d582604c589b","expireAt":1775126363538}

package/dist/.next/cache/.rscinfo

@@ -1 +1 @@
-{"encryption.key":"O1MQTVzzTqHAp+AmHGieR/YbDsRTV3haJ7igC3+CDKc=","encryption.expire_at":1773770546582}
+{"encryption.key":"ht9K53UtRRttTiNu7t5Oj65n4lCPbGM1sV7diyI4Lcs=","encryption.expire_at":1775126363195}