Agent skill

See

使用外部视觉服务或本地 OCR 分析本地图片、截图、视频及明确提供的媒体 URL，输出 Markdown。用户明确要求 see，或当前主模型缺少完成媒体分析所需的视觉能力时使用。不拦截普通文本、任意 URL 或已有原生视觉可以直接完成的任务；不用于生图、编辑图片或自动操作屏幕。使用外部服务前说明发送的媒体与服务范围，无法识别时如实说明限制。

MITAuto-check: notesDocuments & Office

Install See

skills CLI

$ npx skills add oil-oil/see-skill --skill see -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI

$ gh skill install oil-oil/see-skill see --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy

$ git clone --depth 1 https://github.com/oil-oil/see-skill.git skills-src && mkdir -p .claude/skills && cp -r skills-src/see .claude/skills/see && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Paste into Claude Code

Install the "see" agent skill from https://github.com/oil-oil/see-skill/tree/main/see into .claude/skills/see/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "see", then confirm the skill loads.

Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

$skill-installer

$skill-installer install https://github.com/oil-oil/see-skill/tree/main/see

Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.

skills CLI

$ npx skills add oil-oil/see-skill --skill see -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install oil-oil/see-skill see --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/oil-oil/see-skill.git skills-src && mkdir -p .agents/skills && cp -r skills-src/see .agents/skills/see && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "see" agent skill from https://github.com/oil-oil/see-skill/tree/main/see into .agents/skills/see/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "see", then confirm the skill loads.

Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add oil-oil/see-skill --skill see -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install oil-oil/see-skill see --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/oil-oil/see-skill.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/see .cursor/skills/see && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "see" agent skill from https://github.com/oil-oil/see-skill/tree/main/see into .cursor/skills/see/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "see", then confirm the skill loads.

Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

gemini skills install

$ gemini skills install https://github.com/oil-oil/see-skill.git --path see

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add oil-oil/see-skill --skill see -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install oil-oil/see-skill see --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/oil-oil/see-skill.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/see .gemini/skills/see && rm -rf skills-src

Use ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.

Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/

Paste into Gemini CLI

Install the "see" agent skill from https://github.com/oil-oil/see-skill/tree/main/see into .gemini/skills/see/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "see", then confirm the skill loads.

Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

GitHub CLI

$ gh skill install oil-oil/see-skill see

Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).

skills CLI

$ npx skills add oil-oil/see-skill --skill see -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/oil-oil/see-skill.git skills-src && mkdir -p .github/skills && cp -r skills-src/see .github/skills/see && rm -rf skills-src

Use ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.

GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/

Paste into GitHub Copilot

Install the "see" agent skill from https://github.com/oil-oil/see-skill/tree/main/see into .github/skills/see/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "see", then confirm the skill loads.

GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add oil-oil/see-skill --skill see -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install oil-oil/see-skill see --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/oil-oil/see-skill.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/see .opencode/skills/see && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "see" agent skill from https://github.com/oil-oil/see-skill/tree/main/see into .opencode/skills/see/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "see", then confirm the skill loads.

OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

Facts

Skill name: see
GitHub stars: 160
Token cost: ~632 tokens
SKILL.md length: 119 words
Files: 44 (incl. scripts, references)
Skills in repo: 1
Repo updated: Sep 10, 2026
First seen: Oct 7, 2026
Licence: MIT

At a glance

Documents & Office work in your project
SKILL.md covers API Key 配置入口 and 凭据与数据边界
Runs JavaScript scripts from its folder; calls python3; needs ZENMUX_API_KEY and DASHSCOPE_API_KEY

What it does

See is an agent skill from oil-oil/see-skill. 使用外部视觉服务或本地 OCR 分析本地图片、截图、视频及明确提供的媒体 URL，输出 Markdown。用户明确要求 see，或当前主模型缺少完成媒体分析所需的视觉能力时使用。不拦截普通文本、任意 URL 或已有原生视觉可以直接完成的任务；不用于生图、编辑图片或自动操作屏幕。使用外部服务前说明发送的媒体与服务范围，无法识别时如实说明限制。

Its SKILL.md is about 630 tokens, which your agent loads only when the skill is triggered. The skill folder holds 49 other files, including scripts and reference files (for example `agents/openai.yaml`, `references/api-key-setup.md` and `scripts/credential-ui/README.md`).

It sits in Documents & Office. It works with OpenRouter, Linux and macOS. The repository describes itself as: 分析图片、截图和视频，输出可阅读的媒体报告，并在适用时使用本地文字识别。 The licence is MIT.

When your agent uses it

Documents & Office work in your project

Example prompts

“/see”

Requirements

Python 3
Node.js
A credential in ZENMUX_API_KEY
A credential in DASHSCOPE_API_KEY

What it can do on your machine

Read from SKILL.md and the folder at commit 2b7aee8. It shows what the files ask for, not the result of running them.

Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Ships 12 files in scripts/ (JavaScript, from the files we listed), which the agent can run.
Shell commands in SKILL.md call:
- python3
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names these keys or tokens, usually read from environment variables:
- ZENMUX_API_KEY
- DASHSCOPE_API_KEY
- OPENROUTER_API_KEY
- TOKENDANCE_API_KEY
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

See loads about 632 tokens when it runs, and up to ~1.6k if it reads all its reference files. Until then it costs about 44 tokens; SKILL.md has 119 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~44
When it runs · the whole SKILL.md, loaded when a task matches: ~632
With references · SKILL.md plus every file in references/, read only if the agent opens them: ~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

NoteMentions a .env fileSKILL.md:49

KEY`、`TOKENDANCE_API_KEY`。配置读取顺序为环境变量 → `.env.local` → 用户私有配置。

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from oil-oil/see-skill at commit 2b7aee8, republished under its MIT licence (© oil-oil). 119 words, ~632 tokens.

Download SKILL.mdSave it as .claude/skills/see/SKILL.md (or your agent's skills folder). This skill also uses 43 other files; get the full folder from GitHub.

name: see
description: 使用外部视觉服务或本地 OCR 分析本地图片、截图、视频及明确提供的媒体 URL，输出 Markdown。用户明确要求 see，或当前主模型缺少完成媒体分析所需的视觉能力时使用。不拦截普通文本、任意 URL 或已有原生视觉可以直接完成的任务；不用于生图、编辑图片或自动操作屏幕。使用外部服务前说明发送的媒体与服务范围，无法识别时如实说明限制。

See

先检查已有附件、可读媒体路径和当前宿主的视觉能力。原生视觉足够时直接完成；需要外部视觉时，说明将发送哪些媒体给哪个服务，并继承当前任务已经给出的授权。缺少可读输入时才索取路径或 URL。

将当前 Skill 目录解析为绝对路径后调用 scripts/see.sh。外部服务或 OCR 均不可用时如实说明，不隐瞒限制，不把 OCR 文字结果称为完整视觉理解。

安装 Skill 不会更换当前主模型；界面继续显示原来的文本模型是正常的。拖拽或粘贴图片若被拒绝，说明附件在 Skill 启动前已被拦截。不要反复尝试直接读取附件。

首次使用，或用户反馈模型说了「不支持视觉」却没有调用 see 时，先运行：

bash

python3 scripts/onboard.py --status

仅用户明确要求修改全局规则时，才使用 onboard.py --install-agents；它是可选的宿主适配，不是媒体分析的前置步骤。缺少配置时由安全入口收集凭据，不让用户把密钥发到对话。

bash

# 单图
scripts/see.sh image.png

# 视频
scripts/see.sh video.mp4

# 多图并行
scripts/see.sh a.png b.png c.png

# 多图比较或联合判断
scripts/see.sh --together before.png after.png --task "比较界面变化"

# 可选关注点
scripts/see.sh screenshot.png --task "重点识别界面文字"

成功后读取 stdout 中 output_path=<path> 指向的 Markdown。

脚本自动完成：识别图片或视频 → 选择供应商 → 失败时切换供应商。图片无云端时降级到本地视觉；视频自动压缩后原生输入模型，不自行抽帧。多文件默认并行。

图片原图直传。视频按所选服务的模型配置处理；自动保留清晰度、音频和完整时间线。--task 原样发送；没有特殊问题时不要添加。

新配置按上述 API Key 配置说明打开本机页面，由用户亲自保存；选择的配置名、业务 --provider 必须对应。通过 run 包装云端分析命令，不能绕过包装器读取页面凭据。原 onboard 只用于兼容配置、状态或用户明确选择的终端方式。

供应商：zenmux、bailian、openrouter、tokendance、local。图片默认 Qwen3.7 Plus；视频在 ZenMux/OpenRouter 默认 Gemini 3.1 Flash-Lite，其余平台默认 Qwen3.7 Plus。覆盖视频模型用 SEE_VIDEO_MODEL。

也兼容厂商变量：ZENMUX_API_KEY、DASHSCOPE_API_KEY、OPENROUTER_API_KEY、TOKENDANCE_API_KEY。配置读取顺序为环境变量 → .env.local → 用户私有配置。

Windows 私有配置位于 %APPDATA%\see\config.env；macOS/Linux 位于 ~/.config/see/config.env。配置文件权限仅限当前用户，不得复制进 Skill 或项目仓库。

本地降级：

macOS：系统 Vision OCR；有 Swift 时增加场景/人物/人脸/条码/图形结构 → Tesseract
Windows：Windows OCR → Tesseract
Linux：Tesseract

本地后端报错时先运行 python3 scripts/onboard.py --status。macOS 10.15+ 不需要 Xcode；Windows 需要安装系统 OCR 语言；Linux 需要安装 Tesseract。

可选参数只在需要时使用：--together、--provider、--model、--task、--jobs、--ocr-backend。本地视觉结果不等同于多模态模型的完整语义理解。

视频需要任一云端 Key；同一个 Key 同时用于图片和视频。主 Agent 只传路径并读取 output_path，不要自行调用 ffmpeg、抽帧或上传。

API Key 配置入口

需要外部服务凭据时先读API Key 配置与业务读取：复用已有安全入口；本机缺少 Key 时使用随附固定页面，保存后通过业务包装入口读取。内置能力与纯本地流程不要求配置 Key。

凭据与数据边界

云端分析会把选定媒体发给所选供应商，可能计费；本地 OCR 不等于完整视觉理解。新密钥通过固定本机页面存入系统凭据库，普通配置只保存引用。先用任务 Python 环境安装 see/scripts/requirements-credentials.txt（Skill 内相对路径为 scripts/requirements-credentials.txt）。不把 Key 发到聊天、命令参数或日志。旧明文配置不会自动迁移或覆盖：入口会先停止，提示备份后重新配置。状态检查只确认来源，API 验证成功才代表可用。全局指令写入仅在明确调用 --install-agents 时发生。

Files

SKILL.md and 43 other files (scripts, references) in see of oil-oil/see-skill.

SKILL.md
agents/openai.yaml
references/api-key-setup.md
scripts/credential-ui/README.md
scripts/credential-ui/component.json
scripts/credential-ui/manifests/bailian.json
scripts/credential-ui/manifests/default.json
scripts/credential-ui/manifests/openrouter.json
scripts/credential-ui/manifests/profiles.json
scripts/credential-ui/manifests/tokendance.json
scripts/credential-ui/package-lock.json
scripts/credential-ui/package.json
scripts/credential-ui/public/app.js
scripts/credential-ui/public/favicon.svg
scripts/credential-ui/public/index.html
… and 29 more

Open the folder on GitHubat commit 2b7aee8

Compare with similar skills

See next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

See compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
See this skilloil-oil/see-skill	160	—	~632	Automated safety check: Notes	MIT	1 mo ago
Scanned PDF To Epublonghaiqwe/scanned-pdf-to-epub	161	—	~1.9k	Automated safety check: Pass	MIT	26 days ago
Editaplotliushunqi8-hash/editaplot2026	128	—	~5.8k	Automated safety check: Pass	Apache-2.0	20 days ago
Rename Pdfsrealspqrk/autorename-pdf	122	—	~541	Automated safety check: Pass	MIT	1 mo ago
Prismer NotionPrismer-AI/PrismerCloud	1.6k	3 repos	~4.6k	Automated safety check: Pass	MIT	today
File Documentsiamlukethedev/Herald-OS	408	—	~1.7k	Automated safety check: Pass	MIT	yesterday

Similar skills

Scanned PDF To Epub
longhaiqwe/scanned-pdf-to-epub
A skill your agent uses when scanned Chinese PDFs or image-only books need text EPUB conversion on macOS, Windows, or Linux, especially when OCR text, original underlined proper names, 专名号, or…
161 GitHub stars~1.9k tokensUpdated 26 days ago
Documents & OfficeAuto-check passed
Editaplot
liushunqi8-hash/editaplot2026
Analyze local scientific CSV, TXT, XLS, or XLSX data; recommend publication-informed charts and Chinese scientific palettes; freeze a reproducible plan; and automate editable figures through a…
128 GitHub stars~5.8k tokensUpdated 20 days ago
Documents & OfficeAuto-check passed
Rename Pdfs
realspqrk/autorename-pdf
[Dev] Rename PDF files using AI (Python, requires venv). An agent skill from realspqrk/autorename-pdf.
122 GitHub stars~541 tokensUpdated 1 mo ago
Documents & OfficeAuto-check passed
Prismer Notion
Prismer-AI/PrismerCloud
Notion API + ntn CLI: pages, databases, markdown, Workers. An agent skill from Prismer-AI/PrismerCloud.
1.6k GitHub starsUsed in 3 repos~4.6k tokens
Documents & OfficeAuto-check passed
File Documents
iamlukethedev/Herald-OS
Find documents such as invoices among badly named files, read them, name them clearly and file them where the person keeps them, with a plan they approve and an undo
408 GitHub stars~1.7k tokensUpdated yesterday
Documents & OfficeAuto-check passed
Ppt Compressor
LeoYeAI/openclaw-master-skills
This skill should be used when the user wants to compress a PowerPoint (.pptx) file by reducing the size of embedded videos and large images.
2.2k GitHub stars~3.9k tokensUpdated 2 mo ago
Documents & OfficeAuto-check passed

Works with

Questions about See

What does See do?

When should I use See?

See fits situations like: documents & Office work in your project.

How do I install See in Claude Code?

Run `npx skills add oil-oil/see-skill --skill see -a claude-code`. Or copy the skill folder (see in oil-oil/see-skill) into .claude/skills/see in your project. Claude Code loads it when a task matches its description.

How do I install See in Codex?

Run `npx skills add oil-oil/see-skill --skill see -a codex`. Or copy the skill folder (see in oil-oil/see-skill) into .agents/skills/see in your project. Codex loads it when a task matches its description.

Can I use See in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add oil-oil/see-skill --skill see -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/see, .gemini/skills/see, .github/skills/see and .opencode/skills/see in your project.

What does See need to run?

Going by SKILL.md and its folder, See needs JavaScript for the scripts in its folder, the command-line tools its instructions call (python3) and credentials named ZENMUX_API_KEY, DASHSCOPE_API_KEY, OPENROUTER_API_KEY and TOKENDANCE_API_KEY. Our summary lists: Python 3; Node.js; A credential in ZENMUX_API_KEY; A credential in DASHSCOPE_API_KEY.

Does See access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is See safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does See use?

See is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does See use?

About 632 tokens (SKILL.md is roughly 2.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 1k tokens, read only when the agent opens those files.

What are the alternatives to See?

Skills that share tags, products or a category with See: Scanned PDF To Epub (longhaiqwe/scanned-pdf-to-epub, 161 stars), Editaplot (liushunqi8-hash/editaplot2026, 128 stars), Rename Pdfs (realspqrk/autorename-pdf, 122 stars) and Prismer Notion (Prismer-AI/PrismerCloud, 1.6k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains See?

oil-oil (a GitHub user) maintains it in oil-oil/see-skill, which has 160 GitHub stars. The repository was last updated on September 10, 2026.

Source: oil-oil/see-skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.

Install See

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

See

API Key 配置入口

凭据与数据边界

Files

Compare with similar skills

Similar skills

Scanned PDF To Epub

Editaplot

Rename Pdfs

Prismer Notion

File Documents

Ppt Compressor

Related

Questions about See