Claude Approval — Design skill for Claude Code
Approve Claude Code tool calls from your phone via Feishu (Lark) - a zero-dependency PreToolUse hook with fail-closed design.
How to install Claude Approval
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open z1y1n/claude-approval and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Claude Approval does
Approve Claude Code tool calls from your phone via Feishu (Lark) - a zero-dependency PreToolUse hook with fail-closed design.
Alternatives in Design
- Brand Guidelines — Apply Anthropic's official brand colors and typography to artifacts 94.1k ★
- Od Contribute — Open a first-contribution PR (or bug issue) on nexu-io/open-design — works for non-coders too 91.8k ★
- Export Diagram — Export a diagram-design HTML file to .svg and .png next to the source 27.2k ★
README
claude-approval
**Approve Claude Code's tool calls from your phone.**
A small, zero-dependency [Claude Code](https://claude.com/claude-code) `PreToolUse` hook that pushes every tool call to your phone via Feishu (Lark) and blocks until you answer — one tap on **✅ Allow / ⛔ Deny**, or a one-character reply. Walk away from the keyboard without either stalling the agent or handing it unrestricted access to your files.
[中文说明见下方](#中文说明)
The problem
AI coding agents stop and ask for permission before most tool calls. When you're at the keyboard that's one click. When you're not, you have two bad options:
- Leave it waiting — the task stalls until you get back, sometimes on the very first file write.
- Turn on "skip permissions" — the agent now edits and runs whatever it wants, unattended.
This middleware adds a third option: a human-in-the-loop gate that follows you onto your phone.
How it works
Claude Code wants to run a tool
│
▼
PreToolUse hook (hook.cjs)
│
├─ away mode off ──────────────────► exit 0 (identical to not being installed)
├─ read-only tool ─────────────────► exit 0 (Read / Glob / Grep …)
├─ approved in the last 10 min ────► allow (no local permission prompt)
│
▼
send a card to Feishu as the bot
│
│ your phone: 🔐 Claude Code needs your approval
│ ⚡ Run command: npm test
│ [ ✅ Allow ] [ ⛔ Deny ]
│ │
│ └── tap ──► card.action.trigger (WebSocket, no public URL)
│ │
│ ▼
│ button-daemon.cjs (background)
│ verifies the clicker is you,
│ then atomically writes
│ runtime/decisions/.json
│
Related Skills
Distill Decision Skill
A Claude Code Routine that distills a Feishu/Lark user's decision-making style into a reusable skill
Diverge
Run parallel divergent ideation on an open-ended design problem before /plan. Spawns 6 frame generators + crit
SVG Design System
svg-design-system 是一个同时支持 Codex 和 Claude Code 的 SVG 图表设计系统 Skill。它面向文章可视化、产品分析、业务流程、系统架构、项目汇报和飞书/Lark 文档画板场景,帮
Contract Engineering
Typed cross-boundary contract discipline - schema design where illegal states are unrepresentable, versioning
Craft Skill
Frontend design skill for Claude Code & OpenCode: interfaces that look designed, not generated. Anti-cliché ca
Mortiflix OSS
A motion design studio on your own machine: Claude makes the video step by step, you approve every stage. Brin
Related Agents
Engineering Feishu Integration Developer
Full-stack integration expert specializing in the Feishu (Lark) Open Platform — proficient in Feishu bots, min
Smith Inspector
Schema and invariant validator for AgentSmith artifacts. Fail-closed. Invoke whenever a draft agent/skill/comm
A10 Security
A10 SEC — Security gate. Runs SAST/secrets/dependency/IaC checks, threat-models changes, and issues the signed