用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/skilld-dev/skilld --skill skilld命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | skilld |
| description | Search, run, view, install, update, verify, and remove Skills with skilld CLI, including private repository access. |
Use skilld CLI to search for, run, and install Skills.
Run a Skill first. Install a Skill only when the user asks to keep it.
Run a focused search:
skilld search <query> --json
Read data.items before choosing a Skill.
Use each item's selector for a Skill run.
Refine the query when several Skills cover different tasks.
Always use --json when an Agent runs Skill search.
Check the exit code before reading stdout.
If search fails, read the tagged JSON error from stderr.
Use --plain only when another command needs stable text.
Never parse formatted terminal output.
Run the selector returned by search:
skilld run <selector> --json
The command prints SKILL.md and writes no Skill files.
It retains no remote Skill files after the command exits.
Read the printed SKILL.md, then follow it for the current task.
Prefer skilld run for a one-off task.
Read data.files for each supporting file's path, kind, and size.
The initial load prints no supporting file content.
Read one only when the instructions name it:
skilld run <selector> --revision <data.revision> --file <path> --json
Run the exact data.files[].readArgv array when possible.
It contains the source, exact revision, file path, and --json.
Repeat --file to read several files in one command.
Check data.files[].readable before you ask for a file.
A file with readable: false never prints.
Its kind is executable or binary.
Tell the user the Skill needs an install to use that file.
Report which Skill you ran and that skilld wrote no Skill files.
Read data.sourceStatus, data.origin, and data.revision.
A verified status covers where the Skill came from.
It does not cover what the instructions ask you to do.
If the status is unverified, tell the user before you follow the Skill.
Install a Skill when the user wants it in every session. Install a Skill when it must run its own script. Ask the user before you install. An install writes files they did not request.
Install the selector returned by search into the detected Agent target:
skilld install <selector>
Install into global Agent targets:
skilld install <selector> --global
Install this skilld-maintained Skill globally:
skilld install skilld --global
Always use the source selector shown by skilld search.
If private repository access is required, run skilld auth login.
Do not print access tokens or copy them into project files.
After installation, report the installed Skill name and Agent target.
skilld list
skilld view <skill>
Use list to show installed Skills.
Use view to show one Skill's path, source status, and Agent targets.
skilld update <skill>
skilld update --check --json
skilld verify <skill>
skilld remove <skill>
Use update --check --json to inspect update relations without changing files.
Read each data.items[].relation._tag before changing files.
Use update <skill> only when the relation is available.
Treat current, pinned, and notTracked as no action.
If the relation is behind or diverged, ask before changing files.
If the relation is unavailable, report failure.code and failure.message.
Treat unavailable as unknown. Do not infer a newer commit.
Use verify to check the installed bytes and source status.
Use remove only when the request names the Skill to remove.