llamactl 0.3.9__tar.gz → 0.3.10__tar.gz

{llamactl-0.3.9 → llamactl-0.3.10}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.3
 Name: llamactl
-Version: 0.3.9
+Version: 0.3.10
 Summary: A command-line interface for managing LlamaDeploy projects and deployments
 Author: Adrian Lyjak
 Author-email: Adrian Lyjak <adrianlyjak@gmail.com>
 License: MIT
-Requires-Dist: llama-deploy-core[client]>=0.3.9,<0.4.0
-Requires-Dist: llama-deploy-appserver>=0.3.9,<0.4.0
+Requires-Dist: llama-deploy-core[client]>=0.3.10,<0.4.0
+Requires-Dist: llama-deploy-appserver>=0.3.10,<0.4.0
 Requires-Dist: httpx>=0.24.0,<1.0.0
 Requires-Dist: rich>=13.0.0
 Requires-Dist: questionary>=2.0.0

{llamactl-0.3.9 → llamactl-0.3.10}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "llamactl"
-version = "0.3.9"
+version = "0.3.10"
 description = "A command-line interface for managing LlamaDeploy projects and deployments"
 readme = "README.md"
 license = { text = "MIT" }
@@ -9,8 +9,8 @@ authors = [
 ]
 requires-python = ">=3.11, <4"
 dependencies = [
-    "llama-deploy-core[client]>=0.3.9,<0.4.0",
-    "llama-deploy-appserver>=0.3.9,<0.4.0",
+    "llama-deploy-core[client]>=0.3.10,<0.4.0",
+    "llama-deploy-appserver>=0.3.10,<0.4.0",
     "httpx>=0.24.0,<1.0.0",
     "rich>=13.0.0",
     "questionary>=2.0.0",

{llamactl-0.3.9 → llamactl-0.3.10}/src/llama_deploy/cli/commands/init.py

@@ -55,8 +55,6 @@ def init(
 
 def _create(template: str | None, dir: Path | None, force: bool) -> None:
     # defer loading to improve cli startup time
-    from vibe_llama.scaffold import create_scaffold
-    from vibe_llama.scaffold.scaffold import ProjectName
     from vibe_llama.sdk import VibeLlamaStarter
 
     @dataclass
@@ -64,13 +62,9 @@ def _create(template: str | None, dir: Path | None, force: bool) -> None:
         id: str
         name: str
         description: str
-        source: VibeLlamaTemplate | GithubTemplateRepo
+        source: GithubTemplateRepo
         llama_cloud: bool
 
-    @dataclass
-    class VibeLlamaTemplate:
-        name: ProjectName
-
     @dataclass
     class GithubTemplateRepo:
         url: str
@@ -110,42 +104,54 @@ def _create(template: str | None, dir: Path | None, force: bool) -> None:
             id="basic",
             name="Basic Workflow",
             description="A base example that showcases usage patterns for workflows",
-            source=VibeLlamaTemplate(name="basic"),
+            source=GithubTemplateRepo(
+                url="https://github.com/run-llama/template-workflow-basic"
+            ),
             llama_cloud=False,
         ),
         TemplateOption(
             id="document_parsing",
             name="Document Parser",
             description="A workflow that, using LlamaParse, parses unstructured documents and returns their raw text content",
-            source=VibeLlamaTemplate(name="document_parsing"),
+            source=GithubTemplateRepo(
+                url="https://github.com/run-llama/template-workflow-document-parsing"
+            ),
             llama_cloud=True,
         ),
         TemplateOption(
             id="human_in_the_loop",
             name="Human in the Loop",
             description="A workflow showcasing how to use human in the loop with LlamaIndex workflows",
-            source=VibeLlamaTemplate(name="human_in_the_loop"),
+            source=GithubTemplateRepo(
+                url="https://github.com/run-llama/template-workflow-human-in-the-loop"
+            ),
             llama_cloud=False,
         ),
         TemplateOption(
             id="invoice_extraction",
             name="Invoice Extraction",
             description="A workflow that, given an invoice, extracts several key details using LlamaExtract",
-            source=VibeLlamaTemplate(name="invoice_extraction"),
+            source=GithubTemplateRepo(
+                url="https://github.com/run-llama/template-workflow-invoice-extraction"
+            ),
             llama_cloud=True,
         ),
         TemplateOption(
             id="rag",
             name="RAG",
             description="A workflow that embeds, indexes and queries your documents on the fly, providing you with a simple RAG pipeline",
-            source=VibeLlamaTemplate(name="rag"),
+            source=GithubTemplateRepo(
+                url="https://github.com/run-llama/template-workflow-rag"
+            ),
             llama_cloud=False,
         ),
         TemplateOption(
             id="web_scraping",
             name="Web Scraping",
             description="A workflow that, given several urls, scrapes and summarizes their content using Google's Gemini API",
-            source=VibeLlamaTemplate(name="web_scraping"),
+            source=GithubTemplateRepo(
+                url="https://github.com/run-llama/template-workflow-web-scraping"
+            ),
             llama_cloud=False,
         ),
     ]
@@ -202,14 +208,11 @@ def _create(template: str | None, dir: Path | None, force: bool) -> None:
         else:
             shutil.rmtree(dir, ignore_errors=True)
 
-    if isinstance(resolved_template.source, GithubTemplateRepo):
-        copier.run_copy(
-            resolved_template.source.url,
-            dir,
-            quiet=True,
-        )
-    else:
-        asyncio.run(create_scaffold(resolved_template.source.name, str(dir)))
+    copier.run_copy(
+        resolved_template.source.url,
+        dir,
+        quiet=True,
+    )
     # Initialize git repository if git is available
     has_git = False
     try:

{llamactl-0.3.9 → llamactl-0.3.10}/src/llama_deploy/cli/commands/serve.py

@@ -1,3 +1,4 @@
+import asyncio
 import logging
 import os
 from pathlib import Path
@@ -5,15 +6,19 @@ from typing import Literal
 
 import click
 import questionary
+from click.exceptions import Abort, Exit
 from llama_deploy.cli.commands.auth import validate_authenticated_profile
 from llama_deploy.cli.config.env_service import service
 from llama_deploy.cli.config.schema import Auth
 from llama_deploy.cli.options import interactive_option
 from llama_deploy.cli.styles import WARNING
+from llama_deploy.cli.utils.redact import redact_api_key
+from llama_deploy.core.client.manage_client import ControlPlaneClient
 from llama_deploy.core.config import DEFAULT_DEPLOYMENT_FILE_PATH
 from llama_deploy.core.deployment_config import (
     read_deployment_config_from_git_root_or_cwd,
 )
+from llama_deploy.core.schema.projects import ProjectSummary
 from rich import print as rprint
 
 from ..app import app
@@ -108,6 +113,7 @@ def serve(
             build=preview,
         )
         deployment_config = get_deployment_config()
+        _print_connection_summary()
         start_server_in_target_venv(
             cwd=Path.cwd(),
             deployment_file=deployment_file,
@@ -127,6 +133,9 @@ def serve(
             else None,
         )
 
+    except (Exit, Abort):
+        raise
+
     except KeyboardInterrupt:
         logger.debug("Shutting down...")
 
@@ -207,6 +216,32 @@ def _maybe_inject_llama_cloud_credentials(
 
     existing = os.environ.get("LLAMA_CLOUD_API_KEY") or vars.get("LLAMA_CLOUD_API_KEY")
     if existing:
+        # If interactive, allow choosing between env var and configured profile
+        if interactive:
+            choice = questionary.select(
+                "LLAMA_CLOUD_API_KEY detected in environment. Which credentials do you want to use?",
+                choices=[
+                    questionary.Choice(
+                        title=f"Use environment variable - {redact_api_key(existing)}",
+                        value="env",
+                    ),
+                    questionary.Choice(title="Use configured profile", value="profile"),
+                ],
+            ).ask()
+            if choice is None:
+                raise Exit(0)
+            if choice == "profile":
+                # Ensure we have an authenticated profile and inject from it
+                authed = validate_authenticated_profile(True)
+                _set_env_vars_from_profile(authed)
+                return
+            # Default to env var path when cancelled or explicitly chosen
+            _set_env_vars_from_env({**os.environ, **vars})
+            # If no project id provided, try to detect and select one using the env API key
+            if not os.environ.get("LLAMA_DEPLOY_PROJECT_ID"):
+                _maybe_select_project_for_env_key()
+            return
+        # Non-interactive: trust current environment variables
         _set_env_vars_from_env({**os.environ, **vars})
         return
 
@@ -243,3 +278,56 @@ def _maybe_inject_llama_cloud_credentials(
     rprint(
         f"[{WARNING}]Warning: LLAMA_CLOUD_API_KEY is not set and no logged-in profile was found. The app may not work.[/]"
     )
+
+
+def _maybe_select_project_for_env_key() -> None:
+    """When using an env API key, ensure LLAMA_DEPLOY_PROJECT_ID is set.
+
+    If more than one project exists, prompt the user to select one.
+    """
+    api_key = os.environ.get("LLAMA_CLOUD_API_KEY")
+    base_url = os.environ.get("LLAMA_CLOUD_BASE_URL", "https://api.cloud.llamaindex.ai")
+    if not api_key:
+        return
+    try:
+
+        async def _run() -> list[ProjectSummary]:
+            async with ControlPlaneClient.ctx(base_url, api_key, None) as client:
+                return await client.list_projects()
+
+        projects = asyncio.run(_run())
+        if not projects:
+            return
+        if len(projects) == 1:
+            os.environ["LLAMA_DEPLOY_PROJECT_ID"] = projects[0].project_id
+            return
+        # Multiple: prompt selection
+        choice = questionary.select(
+            "Select a project",
+            choices=[
+                questionary.Choice(
+                    title=f"{p.project_name} ({p.deployment_count} deployments)",
+                    value=p.project_id,
+                )
+                for p in projects
+            ],
+        ).ask()
+        if choice:
+            os.environ["LLAMA_DEPLOY_PROJECT_ID"] = choice
+    except Exception:
+        # Best-effort; if we fail to list, do nothing
+        pass
+
+
+def _print_connection_summary() -> None:
+    base_url = os.environ.get("LLAMA_CLOUD_BASE_URL")
+    project_id = os.environ.get("LLAMA_DEPLOY_PROJECT_ID")
+    api_key = os.environ.get("LLAMA_CLOUD_API_KEY")
+    if not base_url and not project_id and not api_key:
+        return
+    redacted = redact_api_key(api_key)
+    env_text = base_url or "-"
+    proj_text = project_id or "-"
+    rprint(
+        f"Connecting to environment: [bold]{env_text}[/], project: [bold]{proj_text}[/], api key: [bold]{redacted}[/]"
+    )

{llamactl-0.3.9 → llamactl-0.3.10}/src/llama_deploy/cli/config/auth_service.py

@@ -3,6 +3,7 @@ import asyncio
 from llama_deploy.cli.auth.client import PlatformAuthClient, RefreshMiddleware
 from llama_deploy.cli.config._config import Auth, ConfigManager, Environment
 from llama_deploy.cli.config.schema import DeviceOIDC
+from llama_deploy.cli.utils.redact import redact_api_key
 from llama_deploy.core.client.manage_client import ControlPlaneClient, httpx
 from llama_deploy.core.schema import VersionResponse
 from llama_deploy.core.schema.projects import ProjectSummary
@@ -123,8 +124,4 @@ class AuthService:
 
 def _auto_profile_name_from_token(api_key: str) -> str:
     token = api_key or "token"
-    cleaned = token.replace(" ", "")
-    first = cleaned[:6]
-    last = cleaned[-4:] if len(cleaned) > 10 else cleaned[-2:]
-    base = f"{first}****{last}"
-    return base
+    return redact_api_key(token)

llamactl-0.3.10/src/llama_deploy/cli/utils/redact.py

@@ -0,0 +1,29 @@
+from __future__ import annotations
+
+
+def redact_api_key(
+    token: str | None,
+    visible_prefix: int = 6,
+    visible_suffix_long: int = 4,
+    visible_suffix_short: int = 2,
+    long_threshold: int = 10,
+    mask: str = "****",
+) -> str:
+    """Redact an API key for display.
+
+    Shows a prefix and suffix with a mask in the middle. If token is short,
+    reduces the suffix length to keep at least two trailing characters visible.
+
+    This mirrors the masking behavior used for profile names.
+    """
+    if not token:
+        return "-"
+    cleaned = token.replace(" ", "")
+    if len(cleaned) <= 0:
+        return "-"
+    first = cleaned[:visible_prefix]
+    last_len = (
+        visible_suffix_long if len(cleaned) > long_threshold else visible_suffix_short
+    )
+    last = cleaned[-last_len:] if last_len > 0 else ""
+    return f"{first}{mask}{last}"