Best use case
approval is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Imported skill approval from langchain
Teams using approval should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/approval/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How approval Compares
| Feature / Agent | approval | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Imported skill approval from langchain
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
"""Approval widget for HITL - using standard Textual patterns."""
from __future__ import annotations
import asyncio
from typing import Any, ClassVar
from textual import events
from textual.app import ComposeResult
from textual.binding import Binding, BindingType
from textual.containers import Container, Vertical, VerticalScroll
from textual.message import Message
from textual.widgets import Static
from deepagents_cli.widgets.tool_renderers import get_renderer
class ApprovalMenu(Container):
"""Approval menu using standard Textual patterns.
Key design decisions (following mistral-vibe reference):
- Container base class with compose()
- BINDINGS for key handling (not on_key)
- can_focus_children = False to prevent focus theft
- Simple Static widgets for options
- Standard message posting
- Tool-specific widgets via renderer pattern
"""
can_focus = True
can_focus_children = False
# CSS is in app.tcss - no DEFAULT_CSS needed
BINDINGS: ClassVar[list[BindingType]] = [
Binding("up", "move_up", "Up", show=False),
Binding("k", "move_up", "Up", show=False),
Binding("down", "move_down", "Down", show=False),
Binding("j", "move_down", "Down", show=False),
Binding("enter", "select", "Select", show=False),
Binding("1", "select_approve", "Approve", show=False),
Binding("y", "select_approve", "Approve", show=False),
Binding("2", "select_reject", "Reject", show=False),
Binding("n", "select_reject", "Reject", show=False),
Binding("3", "select_auto", "Auto-approve", show=False),
Binding("a", "select_auto", "Auto-approve", show=False),
]
class Decided(Message):
"""Message sent when user makes a decision."""
def __init__(self, decision: dict[str, str]) -> None:
super().__init__()
self.decision = decision
def __init__(
self,
action_request: dict[str, Any],
assistant_id: str | None = None,
id: str | None = None, # noqa: A002
**kwargs: Any,
) -> None:
super().__init__(id=id or "approval-menu", classes="approval-menu", **kwargs)
self._action_request = action_request
self._assistant_id = assistant_id
self._tool_name = action_request.get("name", "unknown")
self._tool_args = action_request.get("args", {})
self._description = action_request.get("description", "")
self._selected = 0
self._future: asyncio.Future[dict[str, str]] | None = None
self._option_widgets: list[Static] = []
self._tool_info_container: Vertical | None = None
def set_future(self, future: asyncio.Future[dict[str, str]]) -> None:
"""Set the future to resolve when user decides."""
self._future = future
def compose(self) -> ComposeResult:
"""Compose the widget with Static children.
Layout prioritizes options visibility - they appear at the top so users
always see them even in small terminals.
"""
# Title
yield Static(
f">>> {self._tool_name} Requires Approval <<<",
classes="approval-title",
)
# Options container FIRST - always visible at top
with Container(classes="approval-options-container"):
# Options - create 3 Static widgets
for i in range(3):
widget = Static("", classes="approval-option")
self._option_widgets.append(widget)
yield widget
# Help text right after options
yield Static(
"↑/↓ navigate • Enter select • y/n/a quick keys",
classes="approval-help",
)
# Separator between options and tool details
yield Static("─" * 40, classes="approval-separator")
# Tool info in scrollable container BELOW options
with VerticalScroll(classes="tool-info-scroll"):
self._tool_info_container = Vertical(classes="tool-info-container")
yield self._tool_info_container
async def on_mount(self) -> None:
"""Focus self on mount and update tool info."""
await self._update_tool_info()
self._update_options()
self.focus()
async def _update_tool_info(self) -> None:
"""Mount the tool-specific approval widget."""
if not self._tool_info_container:
return
# Get the appropriate renderer for this tool
renderer = get_renderer(self._tool_name)
widget_class, data = renderer.get_approval_widget(self._tool_args)
# Clear existing content and mount new widget
await self._tool_info_container.remove_children()
approval_widget = widget_class(data)
await self._tool_info_container.mount(approval_widget)
def _update_options(self) -> None:
"""Update option widgets based on selection."""
options = [
"1. Approve (y)",
"2. Reject (n)",
"3. Auto-approve all this session (a)",
]
for i, (text, widget) in enumerate(zip(options, self._option_widgets, strict=True)):
cursor = "› " if i == self._selected else " "
widget.update(f"{cursor}{text}")
# Update classes
widget.remove_class("approval-option-selected")
if i == self._selected:
widget.add_class("approval-option-selected")
def action_move_up(self) -> None:
"""Move selection up."""
self._selected = (self._selected - 1) % 3
self._update_options()
def action_move_down(self) -> None:
"""Move selection down."""
self._selected = (self._selected + 1) % 3
self._update_options()
def action_select(self) -> None:
"""Select current option."""
self._handle_selection(self._selected)
def action_select_approve(self) -> None:
"""Select approve option."""
self._selected = 0
self._update_options()
self._handle_selection(0)
def action_select_reject(self) -> None:
"""Select reject option."""
self._selected = 1
self._update_options()
self._handle_selection(1)
def action_select_auto(self) -> None:
"""Select auto-approve option."""
self._selected = 2
self._update_options()
self._handle_selection(2)
def _handle_selection(self, option: int) -> None:
"""Handle the selected option."""
decision_map = {
0: "approve",
1: "reject",
2: "auto_approve_all",
}
decision = {"type": decision_map[option]}
# Resolve the future
if self._future and not self._future.done():
self._future.set_result(decision)
# Post message
self.post_message(self.Decided(decision))
def on_blur(self, event: events.Blur) -> None:
"""Re-focus on blur to keep focus trapped."""
self.call_after_refresh(self.focus)Related Skills
u0532-engineering-human-approval-router
Operate the "Engineering Human Approval Router" capability in production for workflows. Use when mission execution explicitly requires this capability and outcomes must be reproducible, policy-gated, and handoff-ready.
u01789-human-approval-routing-for-remote-team-collaboration
Operate the "Human Approval Routing for remote team collaboration" capability in production for remote team collaboration workflows. Use when mission execution explicitly requires this capability and outcomes must be reproducible, policy-gated, and handoff-ready.
u01784-human-approval-routing-for-multilingual-translation-services
Operate the "Human Approval Routing for multilingual translation services" capability in production for multilingual translation services workflows. Use when mission execution explicitly requires this capability and outcomes must be reproducible, policy-gated, and handoff-ready.
u01689-human-approval-routing-for-education-support-services
Operate the "Human Approval Routing for education support services" capability in production for education support services workflows. Use when mission execution explicitly requires this capability and outcomes must be reproducible, policy-gated, and handoff-ready.
asset-approval
Use to manage co-marketing asset reviews with traceable evidence and SLAs.
approval-workflow
Manages Human-in-the-Loop (HITL) approval workflows for sensitive actions. Use when creating approval requests, processing approved items, or implementing safety controls for autonomous actions.
approval-gate
ワークフローの重要なフェーズ移行前にユーザーの明示的な承認を必要とする承認ゲートの共通フォーマットとパターンを定義
u01688-human-approval-routing-for-education-support-services
Operate the "Human Approval Routing for education support services" capability in production for education support services workflows. Use when mission execution explicitly requires this capability and outcomes must be reproducible, policy-gated, and handoff-ready.
architecture-review-and-approval
Run structured architecture reviews with stakeholders and sign-off.
bgo
Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.
customer-discovery
Find where potential customers discuss problems online and extract their language patterns. Provides starting points for community research, not exhaustive coverage.
create-prd
This skill should be used when the user asks to "创建PRD", "写产品需求文档", "生成PRD", "新建PRD", "create PRD", "write product requirements document", or mentions "产品需求文档", "PRD模板". Automatically generates comprehensive Chinese PRD documents following 2026 best practices.