google-agents-cli 0.6.0__tar.gz → 0.6.1__tar.gz

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: google-agents-cli
-Version: 0.6.0
+Version: 0.6.1
 Summary: Agents CLI — Agent Development Lifecycle toolchain for ADK
 Author-email: Google LLC <agents-cli-dev-pypi@google.com>
 Requires-Python: >=3.11
@@ -55,7 +55,7 @@ Turn your favorite coding assistant into an expert at building and deploying age
 **Agents CLI in Agent Platform** (`agents-cli`) gives your coding agent the skills and commands to build, scale, govern, and optimize enterprise-grade agents — so you don't have to learn every CLI and service yourself.
 
 **Works seamlessly with:**
-[Gemini CLI](https://github.com/google-gemini/gemini-cli) &nbsp;•&nbsp; [Claude Code](https://docs.anthropic.com/en/docs/claude-code) &nbsp;•&nbsp; [Codex](https://github.com/openai/codex) &nbsp;•&nbsp; [Antigravity](https://antigravity.google/) &nbsp;•&nbsp; *and any other coding agent.*
+[Antigravity CLI](https://antigravity.google/) &nbsp;•&nbsp; [Claude Code](https://docs.anthropic.com/en/docs/claude-code) &nbsp;•&nbsp; [Codex](https://github.com/openai/codex) &nbsp;•&nbsp; *and any other coding agent.*
 
 ## Get Started
 
@@ -78,7 +78,7 @@ npx skills add google/agents-cli
 
 ### 2. Open your coding agent
 
-Launch [Gemini CLI](https://github.com/google-gemini/gemini-cli), [Claude Code](https://docs.anthropic.com/en/docs/claude-code), [Codex](https://github.com/openai/codex), or any coding agent you prefer.
+Launch [Antigravity CLI](https://antigravity.google/), [Claude Code](https://docs.anthropic.com/en/docs/claude-code), [Codex](https://github.com/openai/codex), or any coding agent you prefer.
 
 ### 3. Build your first agent
 
@@ -170,7 +170,7 @@ The Google Cloud agent stack that `agents-cli` builds on:
 
 ## FAQ
 
-**Is this an alternative to Gemini CLI, Claude Code, or Codex?**<br>
+**Is this an alternative to Antigravity CLI, Claude Code, or Codex?**<br>
 No. **`agents-cli` is a tool *for* coding agents, not a coding agent itself.** It provides the CLI commands and skills that make your coding agent better at building, evaluating, and deploying ADK agents on Google Cloud.
 
 **How is this different from just using `adk` directly?**<br>

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/README.md

@@ -22,7 +22,7 @@ Turn your favorite coding assistant into an expert at building and deploying age
 **Agents CLI in Agent Platform** (`agents-cli`) gives your coding agent the skills and commands to build, scale, govern, and optimize enterprise-grade agents — so you don't have to learn every CLI and service yourself.
 
 **Works seamlessly with:**
-[Gemini CLI](https://github.com/google-gemini/gemini-cli)  •  [Claude Code](https://docs.anthropic.com/en/docs/claude-code)  •  [Codex](https://github.com/openai/codex)  •  [Antigravity](https://antigravity.google/)  •  *and any other coding agent.*
+[Antigravity CLI](https://antigravity.google/)  •  [Claude Code](https://docs.anthropic.com/en/docs/claude-code)  •  [Codex](https://github.com/openai/codex)  •  *and any other coding agent.*
 
 ## Get Started
 
@@ -45,7 +45,7 @@ npx skills add google/agents-cli
 
 ### 2. Open your coding agent
 
-Launch [Gemini CLI](https://github.com/google-gemini/gemini-cli), [Claude Code](https://docs.anthropic.com/en/docs/claude-code), [Codex](https://github.com/openai/codex), or any coding agent you prefer.
+Launch [Antigravity CLI](https://antigravity.google/), [Claude Code](https://docs.anthropic.com/en/docs/claude-code), [Codex](https://github.com/openai/codex), or any coding agent you prefer.
 
 ### 3. Build your first agent
 
@@ -137,7 +137,7 @@ The Google Cloud agent stack that `agents-cli` builds on:
 
 ## FAQ
 
-**Is this an alternative to Gemini CLI, Claude Code, or Codex?**<br>
+**Is this an alternative to Antigravity CLI, Claude Code, or Codex?**<br>
 No. **`agents-cli` is a tool *for* coding agents, not a coding agent itself.** It provides the CLI commands and skills that make your coding agent better at building, evaluating, and deploying ADK agents on Google Cloud.
 
 **How is this different from just using `adk` directly?**<br>

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "google-agents-cli"
-version = "0.6.0"
+version = "0.6.1"
 description = "Agents CLI — Agent Development Lifecycle toolchain for ADK"
 readme = "README.md"
 license = "Apache-2.0"

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-adk-code/SKILL.md

@@ -12,7 +12,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-deploy/SKILL.md

@@ -12,7 +12,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-deploy/references/agent-runtime.md

@@ -41,9 +41,10 @@ https://{location}-aiplatform.googleapis.com/reasoningEngines/v1/{resource}/api/
 
 `{agent_directory}` is the app name (the project's `agent_directory`, recorded in
 `deployment_metadata.json`). This is the exact URL `deploy` advertises on
-success, `publish` registers with Gemini Enterprise, and `run` constructs for
-`--mode a2a` against an Agent Runtime URL. All three build it from the shared
-`_agent_runtime_a2a` helper.
+success and `run` constructs for `--mode a2a` against an Agent Runtime URL — both
+authenticate with your Google credentials. On Agent Runtime, `publish` registers
+with **ADK** (`:streamQuery` against the reasoning-engine resource name), not
+this card URL.
 
 ## Deploying
 
@@ -74,7 +75,7 @@ Written by `agents-cli deploy` after a successful deployment:
 }
 ```
 
-Used by: subsequent deploys (update vs create), `agents-cli run --url`, and `agents-cli publish` (to construct the A2A card URL). Cloud Run does not use this file.
+Used by: subsequent deploys (update vs create), `agents-cli run --url`, and `agents-cli publish` (reads the runtime ID for the default ADK registration on Agent Runtime, and constructs the A2A card URL only when A2A registration is explicitly chosen). Cloud Run does not use this file.
 
 If deployment times out but the engine was created, manually populate this file with the engine resource ID.
 

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-eval/SKILL.md

@@ -11,7 +11,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-observability/SKILL.md

@@ -13,7 +13,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-publish/SKILL.md

@@ -15,7 +15,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli
@@ -40,29 +40,22 @@ metadata:
 
 ## Registration Modes
 
-### A2A Registration (default)
+### A2A Registration (Cloud Run / GKE)
 
-Every scaffolded agent serves the Agent-to-Agent protocol, so A2A is the default on **every** deployment target — Agent Runtime included. Pass the agent card URL and the command fetches the card and registers it; display name and description default to the card's `name`/`description`.
+Every scaffolded agent serves the Agent-to-Agent protocol. A2A is the default — and only — registration type on **Cloud Run** and **GKE**, which have no reasoning engine, so Gemini Enterprise registers them over A2A. Pass the agent card URL and the command fetches the card and registers it; display name and description default to the card's `name`/`description`.
 
 ```bash
-# A2A on Agent Runtime (card served through the /api passthrough)
-agents-cli publish gemini-enterprise \
-  --agent-card-url "https://us-east1-aiplatform.googleapis.com/reasoningEngines/v1/projects/123456/locations/us-east1/reasoningEngines/789/api/a2a/app/.well-known/agent-card.json" \
-  --gemini-enterprise-app-id projects/123456/locations/global/collections/default_collection/engines/my-app
-
 # A2A on Cloud Run / GKE
 agents-cli publish gemini-enterprise \
   --agent-card-url https://my-service-abc123.us-east1.run.app/a2a/app/.well-known/agent-card.json \
   --gemini-enterprise-app-id projects/123456/locations/global/collections/default_collection/engines/my-app
 ```
 
-Pass `--display-name` / `--description` to override the card defaults. On Agent Runtime, `--agent-card-url` can be omitted when `deployment_metadata.json` is present — see [Auto-Detection from Metadata](#auto-detection-from-metadata).
-
-> **OAuth authorization is not supported yet for A2A on Agent Runtime.** If your agent needs an OAuth authorization (`--authorization-id`), register it as ADK instead (`--registration-type adk`) — Gemini Enterprise invokes it natively via `:streamQuery`, which does support authorizations.
+Pass `--display-name` / `--description` to override the card defaults. For Agent Runtime, use ADK registration (below).
 
-### ADK Registration
+### ADK Registration (default on Agent Runtime)
 
-Opt in with `--registration-type adk` for Agent Runtime agents you want Gemini Enterprise to invoke natively via `:streamQuery` instead of over A2A — also the path to use when the agent needs an OAuth authorization (`--authorization-id`), which A2A on Agent Runtime does not yet support. The agent is registered directly via its reasoning engine resource name.
+This is the **default and recommended registration for Agent Runtime** deployments: Gemini Enterprise invokes the agent natively via `:streamQuery` on its reasoning engine resource, authenticating end-to-end. It's also the path to use when the agent needs an OAuth authorization (`--authorization-id`). The agent is registered directly via its reasoning engine resource name; no agent card URL is needed.
 
 ```bash
 agents-cli publish gemini-enterprise \
@@ -124,9 +117,9 @@ agents-cli publish gemini-enterprise --interactive
 | `--display-name` | `GEMINI_DISPLAY_NAME` | Display name in Gemini Enterprise |
 | `--description` | `GEMINI_DESCRIPTION` | Agent description |
 | `--tool-description` | `GEMINI_TOOL_DESCRIPTION` | Tool description (ADK mode only, defaults to description) |
-| `--registration-type` | `REGISTRATION_TYPE` | `adk` or `a2a` (defaults to `a2a` for scaffolded agents if not set) |
+| `--registration-type` | `REGISTRATION_TYPE` | `adk` or `a2a` (defaults to `adk` on Agent Runtime, `a2a` on Cloud Run / GKE) |
 | `--agent-card-url` | `AGENT_CARD_URL` | Agent card URL for A2A registration |
-| `--deployment-target` | `DEPLOYMENT_TARGET` | `agent_runtime`, `cloud_run`, or `gke` (affects A2A auth method) |
+| `--deployment-target` | `DEPLOYMENT_TARGET` | `agent_runtime`, `cloud_run`, or `gke` (sets the default registration type — ADK on Agent Runtime, A2A on Cloud Run / GKE — and the A2A auth method) |
 | `--project-id` | `GOOGLE_CLOUD_PROJECT` | GCP project ID for billing |
 | `--project-number` | `PROJECT_NUMBER` | GCP project number (used for Gemini Enterprise lookup) |
 | `--authorization-id` | `GEMINI_AUTHORIZATION_ID` | OAuth authorization resource name |
@@ -141,11 +134,10 @@ agents-cli publish gemini-enterprise --interactive
 When `deployment_metadata.json` exists, the command automatically:
 
 - Reads the **agent runtime ID** (`remote_agent_runtime_id`)
-- Determines the **registration type**: defaults to **A2A** on every deployment target, since every scaffolded agent serves A2A. Opt into native ADK invocation (`:streamQuery`) with `--registration-type adk` on Agent Runtime.
-- Constructs the **agent card URL** for A2A registrations on Agent Runtime. The card is served through the Agent Engine `/api` passthrough, so the URL is `https://{location}-aiplatform.googleapis.com/reasoningEngines/v1/{resource}/api/a2a/{agent_directory}/.well-known/agent-card.json` — built from `remote_agent_runtime_id` and `agent_directory` in `deployment_metadata.json` (re-deploy if the metadata predates `agent_directory`).
+- Determines the **registration type**: defaults to **ADK** (native `:streamQuery`) on **Agent Runtime**, and **A2A** on **Cloud Run / GKE** (which have no reasoning engine). Override with `--registration-type`.
 - Determines the **deployment target** for authentication
 
-This means that for the simplest case (an A2A agent on Agent Runtime), you only need to provide the Gemini Enterprise app ID:
+This means that for the simplest case (an agent on Agent Runtime, registered as ADK), you only need to provide the Gemini Enterprise app ID:
 
 ```bash
 agents-cli publish gemini-enterprise \

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-scaffold/SKILL.md

@@ -12,7 +12,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli
@@ -73,7 +73,7 @@ agents-cli scaffold create <project-name> \
 **Constraints:**
 - Project name must be **26 characters or less**, lowercase letters, numbers, and hyphens only.
 - Do NOT `mkdir` the project directory before running `create` — the CLI creates it automatically. If you mkdir first, `create` will fail or behave unexpectedly.
-- Auto-detect the guidance filename based on the IDE you are running in and pass `--agent-guidance-filename` accordingly (`GEMINI.md` for Gemini CLI, `CLAUDE.md` for Claude Code, `AGENTS.md` for OpenAI Codex/other).
+- Auto-detect the guidance filename based on the IDE you are running in and pass `--agent-guidance-filename` accordingly (`AGENTS.md` for Antigravity CLI/OpenAI Codex/other, `CLAUDE.md` for Claude Code, `GEMINI.md` for Gemini CLI).
 - When enhancing an existing project, check where the agent code lives. If it's not in `app/`, pass `--agent-directory <dir>` (e.g. `--agent-directory agent`). Getting this wrong causes enhance to miss or misplace files.
 
 ### Reference Files

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/skills/google-agents-cli-workflow/SKILL.md

@@ -12,7 +12,7 @@ description: >
 metadata:
   author: Google
   license: Apache-2.0
-  version: 0.6.0
+  version: 0.6.1
   requires:
     bins:
       - agents-cli
@@ -23,11 +23,11 @@ metadata:
 
 > **STOP — Do NOT write code yet.** If no project exists, scaffold first with `agents-cli scaffold create <name>`. If the user already has code, use `agents-cli scaffold enhance .` to add the agents-cli structure. Run `agents-cli info` to check if a project already exists. Skipping this leads to missing eval boilerplate, CI/CD config, and project conventions.
 
-**agents-cli** is a CLI and skills toolkit for building, evaluating, and deploying agents on Google Cloud using the [Agent Development Kit (ADK)](https://adk.dev/). It works with any coding agent — Gemini CLI, Claude Code, Codex, or others. Install with `uvx google-agents-cli setup`.
+**agents-cli** is a CLI and skills toolkit for building, evaluating, and deploying agents on Google Cloud using the [Agent Development Kit (ADK)](https://adk.dev/). It works with any coding agent — Antigravity CLI, Claude Code, Codex, or others. Install with `uvx google-agents-cli setup`.
 
 
-> Requires: google-agents-cli ~= 0.6.0
-> If version is behind, run: uv tool install "google-agents-cli~=0.6.0"
+> Requires: google-agents-cli ~= 0.6.1
+> If version is behind, run: uv tool install "google-agents-cli~=0.6.1"
 
 > Check version: agents-cli info
 > [Install uv](https://docs.astral.sh/uv/getting-started/installation/index.md) first if needed.

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/_tools.py

@@ -15,6 +15,7 @@
 """Tool resolution utilities."""
 
 import os
+import re
 import shlex
 import shutil
 import subprocess
@@ -25,6 +26,13 @@ import click
 
 _tool_paths: dict[str, str] = {}
 
+# Matches ANSI escape sequences (CSI/SGR), used to scrub any residual color
+# codes from captured subprocess output. We disable color at the source via
+# NO_COLOR/FORCE_COLOR, but strip defensively in case the tool emits them anyway
+# (e.g. on Windows PowerShell, where embedded escapes have caused error lines
+# to render as the previous line's color — see b/525049570).
+_ANSI_RE = re.compile(r"\x1b\[[0-9;?]*[a-zA-Z]")
+
 _GCLOUD_RELATIVE_PATH = (
     Path("Google") / "Cloud SDK" / "google-cloud-sdk" / "bin" / "gcloud.cmd"
 )
@@ -148,10 +156,17 @@ def run_npx_skills(args: list[str], spinner_msg: str) -> list[str]:
     All non-noise lines are printed immediately. Only concise summary
     lines (e.g. "Installed 6 skills", "Found 6 skills") are collected
     for the end summary.
+
+    Failure lines (e.g. "✗ Failed to update <skill>") are re-colored red and
+    counted; if any are observed the command raises even when npx itself
+    exits 0, since `npx skills update` reports per-skill failures via stdout
+    but still exits cleanly.
+
     Returns:
         A list of summary-worthy lines (short, no decorative content).
     Raises:
-        click.ClickException: If the npx process exits non-zero.
+        click.ClickException: If the npx process exits non-zero, or if any
+            per-skill failures were observed in the streamed output.
     """
     from google.agents.cli._runner import popen_resolved
     from google.agents.cli._skills_check import SKILLS_NPX_PACKAGE
@@ -159,7 +174,16 @@ def run_npx_skills(args: list[str], spinner_msg: str) -> list[str]:
     full_args = ["npx", "-y", SKILLS_NPX_PACKAGE, *args]
     click.secho(f"  \u25b8 {shlex.join(full_args)}", fg="cyan", dim=True)
 
-    summary_lines = []
+    # Disable color in the child process. We re-emit each captured line via
+    # click.echo, so any ANSI sequences from the child would either render as
+    # literal escape codes or, on Windows PowerShell, bleed across line
+    # boundaries — causing error lines to inherit the previous line's color
+    # (e.g. green progress markers making "Failed to update" appear green).
+    # See b/525049570.
+    child_env = {**os.environ, "NO_COLOR": "1", "FORCE_COLOR": "0"}
+
+    summary_lines: list[str] = []
+    failure_lines: list[str] = []
     # Leave stderr attached to the terminal (stderr=None) rather than capturing
     # it. We only ever echo stderr, so piping it would just add a second stream
     # we'd have to drain concurrently to avoid a pipe-buffer deadlock.
@@ -170,12 +194,15 @@ def run_npx_skills(args: list[str], spinner_msg: str) -> list[str]:
         text=True,
         encoding="utf-8",
         errors="replace",
+        env=child_env,
     )
 
     # Read stdout line-by-line
     assert proc.stdout is not None
     for line in proc.stdout:
-        stripped = line.strip()
+        # Defensively strip any leaked ANSI escape sequences so we control how
+        # the line renders.
+        stripped = _ANSI_RE.sub("", line).strip()
         if not stripped:
             continue
         # Skip npx download/cache noise
@@ -184,8 +211,10 @@ def run_npx_skills(args: list[str], spinner_msg: str) -> list[str]:
         # Skip ASCII art banners (block characters)
         if any(ch in stripped for ch in "█╗╔║╚╝"):
             continue
-        # Strip leading box-drawing / bullet prefixes to extract text
-        clean = stripped.lstrip("┌┐└┘├┤│◇●◆✓─╮╯ ")
+        # Strip leading box-drawing / bullet / status-glyph prefixes to extract
+        # text. We intentionally also strip "✗" so we can re-emit the line with
+        # our own (correct) red coloring below.
+        clean = stripped.lstrip("┌┐└┘├┤│◇●◆✓✗─╮╯ ")
         if not clean:
             continue
         # Skip per-agent detail lines
@@ -194,7 +223,17 @@ def run_npx_skills(args: list[str], spinner_msg: str) -> list[str]:
         # Skip purely decorative headers (e.g. "Installation Summary ────")
         if "──" in clean:
             continue
-        click.echo(f"  {clean}")
+
+        # Detect per-skill failures. npx skills reports these as
+        # "✗ Failed to update <skill>" followed by a "Failed to update N
+        # skill(s)" summary line. Either form is a hard failure for us.
+        is_failure = clean.startswith("Failed ") or "Failed to update" in clean
+        if is_failure:
+            failure_lines.append(clean)
+            click.secho(f"  ✗ {clean}", fg="red")
+        else:
+            click.echo(f"  {clean}")
+
         # Collect concise summary-worthy lines for the recap
         if len(clean) < 80 and clean.startswith(
             ("Installed", "Found", "Done", "Removed", "Updated")
@@ -207,4 +246,11 @@ def run_npx_skills(args: list[str], spinner_msg: str) -> list[str]:
         # stderr already streamed to the terminal above.
         raise click.ClickException("npx skills failed")
 
+    if failure_lines:
+        # npx skills exits 0 even when individual skill operations fail, so we
+        # surface those failures as a non-zero exit ourselves.
+        raise click.ClickException(
+            f"npx skills reported {len(failure_lines)} failure(s); see output above."
+        )
+
     return summary_lines

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/publish/cmd_publish.py

@@ -846,12 +846,19 @@ def _match_adk_agent(agent: dict, agent_runtime_id: str) -> bool:
     return re_name == agent_runtime_id
 
 
-def _match_a2a_agent(agent: dict, agent_card_url: str) -> bool:
-    """Whether ``agent`` is the A2A registration for ``agent_card_url``."""
+def _match_a2a_agent(agent: dict, card_url: str | None) -> bool:
+    """Whether ``agent``'s stored card has the same ``url`` as ``card_url``.
+
+    Compares the card's ``url`` (the RPC endpoint, e.g. ``.../a2a/<dir>``), not
+    the ``.well-known/agent-card.json`` fetch URL: GE stores the card body, so
+    matching on the fetch URL never hits and re-publish would duplicate.
+    """
+    if not card_url:
+        return False
     a2a_def = agent.get("a2aAgentDefinition") or agent.get("a2a_agent_definition") or {}
     raw_card = a2a_def.get("jsonAgentCard") or a2a_def.get("json_agent_card") or "{}"
     try:
-        return json.loads(raw_card).get("url") == agent_card_url
+        return json.loads(raw_card).get("url") == card_url
     except json.JSONDecodeError:
         return False
 
@@ -1000,7 +1007,7 @@ def register_a2a_agent(
         base_endpoint=base_endpoint,
         headers=headers,
         payload=payload,
-        match=lambda agent: _match_a2a_agent(agent, agent_card_url),
+        match=lambda agent: _match_a2a_agent(agent, agent_card.get("url")),
     )
 
 
@@ -1173,10 +1180,19 @@ def _list_gemini_enterprise_apps(project_id: str | None, interactive: bool) -> N
     console.print(table)
 
 
-def default_registration_type(*, is_a2a: bool, has_agent_card_url: bool) -> str:
-    """A2A when the agent serves an A2A card, otherwise ADK."""
+def default_registration_type(
+    *, deployment_target: str | None, is_a2a: bool, has_agent_card_url: bool
+) -> str:
+    """Pick the Gemini Enterprise registration path.
+
+    An explicit agent card URL forces A2A. Agent Runtime defaults to native ADK
+    invocation (``:streamQuery``); Cloud Run / GKE have no reasoning engine, so
+    they register over A2A.
+    """
     if has_agent_card_url:
         return "a2a"
+    if deployment_target == "agent_runtime":
+        return "adk"
     if is_a2a:
         return "a2a"
     return "adk"
@@ -1365,6 +1381,10 @@ def register_gemini_enterprise(
     if not resolved_registration_type:
         if provided_agent_card_url or metadata:
             resolved_registration_type = default_registration_type(
+                deployment_target=(
+                    deployment_target
+                    or (metadata.get("deployment_target") if metadata else None)
+                ),
                 is_a2a=metadata.get("is_a2a", False) if metadata else False,
                 has_agent_card_url=bool(provided_agent_card_url),
             )
@@ -1434,6 +1454,15 @@ def register_gemini_enterprise(
                 if metadata
                 else "cloud_run"
             )
+
+        if deployment_target == "agent_runtime":
+            logging.warning(
+                "A2A registration on Agent Runtime is not recommended for "
+                "Gemini Enterprise; prefer ADK registration "
+                "(--registration-type adk), which Gemini Enterprise invokes "
+                "natively via :streamQuery. Proceeding because A2A was requested "
+                "explicitly."
+            )
     else:
         # ADK mode - no agent_card_url needed
         agent_card_url = None

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/scaffold/base_templates/go/README.md

@@ -23,7 +23,7 @@ A Go agent built with Google's Agent Development Kit (ADK).
 └── Makefile             # Common commands
 ```
 
-> **Tip:** Use [Gemini CLI](https://github.com/google-gemini/gemini-cli) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
+> **Tip:** Use [Antigravity CLI](https://antigravity.google/) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
 
 ## Requirements
 

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/scaffold/base_templates/java/README.md

@@ -29,7 +29,7 @@ A Java agent built with Google's Agent Development Kit (ADK).
 └── Makefile                 # Common commands
 ```
 
-> **Tip:** Use [Gemini CLI](https://github.com/google-gemini/gemini-cli) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
+> **Tip:** Use [Antigravity CLI](https://antigravity.google/) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
 
 ## Requirements
 

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/scaffold/base_templates/python/README.md

@@ -27,7 +27,7 @@ Agent generated with `agents-cli` version `{{ cookiecutter.package_version }}`
 └── pyproject.toml             # Project dependencies
 ```
 
-> 💡 **Tip:** Use [Gemini CLI](https://github.com/google-gemini/gemini-cli) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
+> 💡 **Tip:** Use [Antigravity CLI](https://antigravity.google/) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
 
 ## Requirements
 

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/scaffold/base_templates/python/{{cookiecutter.agent_directory}}/app_utils/a2a.py

@@ -61,6 +61,37 @@ def _default_capabilities() -> AgentCapabilities:
         ],
     )
 
+{%- if cookiecutter.deployment_target == 'agent_runtime' %}
+
+
+def _resolve_app_url(app_url: str | None) -> str:
+    """Resolve the public base URL advertised inside the agent card.
+
+    Falls back in order: explicit ``app_url``, the ``APP_URL`` env var, the
+    Agent Runtime ``/api`` passthrough self-built from runtime env vars (valid
+    on the first deploy, before the CLI knows the server-assigned engine ID),
+    then a local default.
+    """
+    if app_url:
+        return app_url
+    if env_url := os.getenv("APP_URL"):
+        return env_url
+
+    agent_engine_id = os.getenv("GOOGLE_CLOUD_AGENT_ENGINE_ID")
+    project = os.getenv("GOOGLE_CLOUD_PROJECT")
+    # Not GOOGLE_CLOUD_LOCATION: the agent pins it to "global", which would build
+    # an invalid "global-aiplatform.googleapis.com" URL.
+    location = os.getenv("GOOGLE_CLOUD_AGENT_ENGINE_LOCATION", "us-east1")
+    if agent_engine_id and project and location:
+        return (
+            f"https://{location}-aiplatform.googleapis.com/reasoningEngines/v1"
+            f"/projects/{project}/locations/{location}"
+            f"/reasoningEngines/{agent_engine_id}/api"
+        )
+
+    return "http://0.0.0.0:8000"
+{%- endif %}
+
 
 async def attach_a2a_routes(
     app: FastAPI,
@@ -82,7 +113,11 @@ async def attach_a2a_routes(
     ``APP_URL``). Call once per app — typically in a FastAPI ``lifespan``, since
     the card is built asynchronously; repeated calls register duplicate routes.
     """
+{%- if cookiecutter.deployment_target == 'agent_runtime' %}
+    resolved_app_url = _resolve_app_url(app_url)
+{%- else %}
     resolved_app_url = app_url or os.getenv("APP_URL", "http://0.0.0.0:8000")
+{%- endif %}
     resolved_agent_version = agent_version or os.getenv("AGENT_VERSION", "0.1.0")
     resolved_capabilities = capabilities or _default_capabilities()
 

{google_agents_cli-0.6.0 → google_agents_cli-0.6.1}/src/google/agents/cli/scaffold/base_templates/typescript/README.md

@@ -27,7 +27,7 @@ This project is organized as follows:
 └── vitest.config.ts     # Vitest test configuration
 ```
 
-> 💡 **Tip:** Use [Gemini CLI](https://github.com/google-gemini/gemini-cli) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
+> 💡 **Tip:** Use [Antigravity CLI](https://antigravity.google/) for AI-assisted development - project context is pre-configured in `GEMINI.md`.
 
 ## Requirements
 
@@ -93,7 +93,7 @@ This template follows a "bring your own agent" approach - you focus on your busi
 4. **Deploy:** Set up and initiate the CI/CD pipelines, customizing tests as necessary. Refer to the [deployment section](#deployment) for comprehensive instructions. For streamlined infrastructure deployment, simply run `uvx google-agents-cli infra cicd`. Currently supports GitHub with both Google Cloud Build and GitHub Actions as CI/CD runners.
 5. **Monitor:** Track performance and gather insights using BigQuery telemetry data, Cloud Logging, and Cloud Trace to iterate on your application.
 
-The project includes a `GEMINI.md` file that provides context for AI tools like Gemini CLI when asking questions about your template.
+The project includes a `GEMINI.md` file that provides context for AI tools like Antigravity CLI when asking questions about your template.
 
 
 ## Deployment