Claude Code Eyes

Look at real hardware, screens, panels, and wiring on a desk or workbench through a camera, so Claude can SEE physical output that no unit test, log, or API can. Use whenever the user says "are you seeing this?", "olha isso", "look at this", "watch this", "can you see...", "check the camera", "look at the screen" / "look at the display", "is this wired right?", or "what's this part?"; asks to visually verify a rendered display, LCD, LED, or panel; wants a wiring, polarity, or voltage-rail check before power-on; or needs to confirm a font, layout, clipping, or on-screen value renders correctly

52 stars 12 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add fcavalcantirj/claude-code-eyes --skill claude-code-eyes

Manual install

git clone --depth 1 https://github.com/fcavalcantirj/claude-code-eyes.git /tmp/claude-code-eyes
mkdir -p ~/.claude/skills
cp -r /tmp/claude-code-eyes/. ~/.claude/skills/claude-code-eyes

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# claude-code-eyes — let Claude look at the real world

Some outputs live where no test, log, or API can reach them: a display panel, an
LED, a breadboard, a rack of wires. This skill grabs the current camera frame and
lets you `Read` it, so you answer from what you actually see — not from what the
code *should* produce.

## How

1. Run: `bash <path-to-skill>/snap.sh` — it prints the path(s) of the captured
   JPEG/PNG. (It works from any directory.)
2. `Read` each printed path — the Read tool renders images visually.
3. Answer from the frame. Reference positions concretely: "the red wire on the
   left rail, third hole down", not "a wire near the top".

If a capture fails, read the error before retrying — it names the actual cause
and the fix. One case is worth knowing: Claude Code runs Bash in a sandbox that
blocks LAN addresses, so a camera your own terminal reaches fine can be
unreachable here. `snap.sh` detects that and prints the fix already filled in —
either approve Claude's offer to rerun the command outside the sandbox, or add the
printed `sandbox.excludedCommands` entry to `~/.claude/settings.json` (the network
allowlist does *not* cover private addresses). Frames are written under `./.claude-code-eyes/` in the working
directory, not `$TMPDIR`, because sandboxed commands and the Read tool resolve
`$TMPDIR` differently. See the README's Troubleshooting section.

**When you cannot read something, zoom instead of guessing.** If a label, marking,

Read the full skill on GitHub

Details

Skill name: claude-code-eyes
Repository: fcavalcantirj/claude-code-eyes
License: MIT
Repo updated: 21 August 2026
Listed in: heilcheng/awesome-agent-skills

FAQ

What does the Claude Code Eyes skill do?

How do I install the Claude Code Eyes skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add fcavalcantirj/claude-code-eyes --skill claude-code-eyes. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Claude Code Eyes popular?

It lives in fcavalcantirj/claude-code-eyes, which has 52 GitHub stars and 12 forks, and is included in 1 curated skill list.

Related skills

Mcdonalds

Lemon177777/mcd-omni-assistant

麦当劳中国官方 MCP 全能助手。当用户提到麦当劳、麦乐送、开心乐园餐、麦麦省优惠券、领券、点麦当劳、外送到家、得来速、麦当劳积分、积分抽奖、积分商城兑换、麦当劳餐品热量/营养、麦当劳订单查询或取消、麦当劳周边玩具与活动时使用。覆盖官方 35 个工具的全部能力，内置领券、到店点餐、外送、玩具雷达、积分抽奖、热量计算、优惠计算器七大工作流。

AI, LLMs & Agents

B0

risingwavelabs/box0

Delegate tasks to AI agents via Box0. Use when the user asks to review code, check security, run tests, compare tools, get multiple perspectives, research a topic, analyze data, write docs, or any task that could benefit from specialized or parallel execution. Also use when the user mentions agent names or says "ask", "delegate", "get opinions from", or "have someone".

AI, LLMs & Agents

Promptmaster

Jakeschincariol/promptmaster-skill

Rewrite the user's raw request into an optimized prompt before answering it, audit it against the 35 credit-waste patterns in waste-patterns.json, and enforce every locked decision in decisions.md so nothing gets re-explained and nothing contradicts itself. Use whenever the user types /promptmaster, prefixes a request with it, or asks to optimize / tighten / fix a prompt.

AI, LLMs & Agents22

Aura

smorky850612/Aurakit

All-in-one fullstack dev engine. /aura: 46 modes (build/fix/clean/deploy/review/spec/lore/ax/experiment/payment/debug/qa/orchestrate/escalate+), 6-layer security with 32 hooks, tiered models (ZERO/ECO/PRO/MAX), 8 languages, 16 specialized agents, SPEC/EARS/TRUST5/XLOOP/RALF/Autopus absorbed. ~55% token savings.

AI, LLMs & AgentsIn 2 lists

Partner Skill

LearnPrompt/partner-skill

搭子.skill / Partner — two-direction cost-split workflow between Claude Code and Codex. Direction A (Codex-driven): Codex orchestrates and implements; Claude Code plans (/goal), polishes UI/interaction, and runs final /codex:review in one reused session. Direction B (Claude-driven): Claude plans and splits the work, runs the idea-king adversarial gate, delegates to Codex background jobs, monitors, and full-reviews before accepting. Use on "搭子" / "双向搭子" / "搭子，恢复" (resume from .partner/), "搭子，配置" (first-run setup wizard), "搭子，试跑" (identity tryout report), 分工给 codex / 让 codex 做 / codex 后台跑 / Claude

AI, LLMs & Agents

Qu AI Wei

baibanbao/qu-ai-wei

王佩专用的去 AI 味系统。触发词：去AI味、去味。合并三家——王佩写作系统 v4.0、humanizer-zh、lieflat-less-ai-tone（含实测数据）。先选模式（自家稿 / 白名单），再过硬边界，再下刀。不判断作者身份，不保证通过 AI 检测器。

AI, LLMs & Agents15

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.