distribute-skill-to-all-agents – skill Agent & Tự động hóa cho agent AI
Phân phối một skill tới các thư mục skill của agent đã cấu hình, đồng thời tôn trọng bố cục symlink cục bộ.
Cài skill distribute-skill-to-all-agents bằng một lệnh
Chọn công cụ bạn đang dùng, copy lệnh và dán vào terminal. Skill sẽ nằm trong thư mục skills của công cụ đó và sẵn sàng dùng ở phiên làm việc tiếp theo.
Claude Code
mkdir -p ~/.claude/skills && curl -fsSL https://skill.dn1nguoi.com/z/distribute-skill-to-all-agents-37d1d.tgz | tar xz -C ~/.claude/skills
Gọi trong Claude Code: /distribute-skill-to-all-agents – rồi mô tả công việc bạn muốn làm.
Codex
mkdir -p ~/.codex/skills && curl -fsSL https://skill.dn1nguoi.com/z/distribute-skill-to-all-agents-37d1d.tgz | tar xz -C ~/.codex/skills
Gọi trong Codex: $distribute-skill-to-all-agents – rồi mô tả công việc bạn muốn làm.
Hermes Agent
mkdir -p ~/.hermes/skills && curl -fsSL https://skill.dn1nguoi.com/z/distribute-skill-to-all-agents-37d1d.tgz | tar xz -C ~/.hermes/skills
Gọi trong Hermes Agent: /distribute-skill-to-all-agents – rồi mô tả công việc bạn muốn làm.
DeepSeek
mkdir -p ~/.deepcode/skills && curl -fsSL https://skill.dn1nguoi.com/z/distribute-skill-to-all-agents-37d1d.tgz | tar xz -C ~/.deepcode/skills
Gọi trong DeepSeek: /distribute-skill-to-all-agents – rồi mô tả công việc bạn muốn làm.
Nội dung SKILL.md
Nội dung SKILL.md gốc bằng tiếng Anh.
Distribute a Skill Across All Agents
When to Use
- Use when a skill should be made available across multiple local agent skill folders.
- Use when the user asks to sync or distribute skill updates to other agents.
Discover actual host locations
The upstream author used ~/.agents/skills, ~/.claude/skills, ~/.pi/agent/skills and ~/.hermes/skills. These are candidate paths, not facts about this user’s machine or discovery precedence. Inspect only authorized configured hosts and their documented skill paths; do not create an absent host or replace a real directory with a symlink merely to match the upstream layout.
Workflow
1. Identify the canonical reviewed source and exact target hosts from the user’s request. Read the entire skill tree and preserve bundled files, attribution and license.
2. Inspect each destination with ls -ld and resolve links. Record source and destination real paths; skip aliases of the same tree and reject unexpected links outside the authorized location.
3. Back up an existing independent target privately. Preview a file-level diff, including deletions and modes; resolve local divergence before overwriting. Validate the skill ID as one directory name and quote paths.
4. Copy only the selected tree to the authorized independent destination. When removal of obsolete files is necessary, preview deletion with a dry run before applying; never use a broad global rsync --delete target.
5. Compare relative file lists and cryptographic content hashes for every file, plus relevant modes and links. Equal SKILL.md byte counts do not prove tree equality or a correct symlink.
6. Verify discovery through each host’s actual mechanism. Report installed bytes separately from a running session reload; reload only the authorized host when required and preserve unrelated sessions.
Updating an existing distributed skill
Repeat the same source/target, divergence, backup and whole-tree verification procedure. Do not propagate host-specific credentials, private memory or unreviewed executable files. A global installation request does not authorize publishing the skill or changing host permission stores.
When NOT to Use This Skill
- Skill is project-specific → put it in
./.claude/skills/,./.pi/agent/skills/, etc. inside the repo, not globally. - Editing one agent's skill only (e.g. a Hermes-only workflow) → patch that file directly, don't propagate.
- Removing a skill globally is destructive. First show the exact skill directories
that would be removed, confirm with the user, then use the user's preferred
safe deletion method for ~/.agents/skills/ and ~/.hermes/skills/.
Limitations
- Adapted from
davidondrej/skills; verify local paths, tools, credentials, and agent features before acting. - For commands, remote access, scheduling, browser automation, or file-changing workflows, get explicit user approval and confirm the target environment first.
Example: distribute one reviewed skill
The user requests the reviewed release-notes skill for their already-configured Claude host. The canonical source is /workspace/team-skills/release-notes; the verified host skill folder is $HOME/.claude/skills.
1. Inspect ls -ld "$HOME/.claude/skills" "$HOME/.claude/skills/release-notes" (an absent skill target is expected for a first install). Resolve the parent and source paths; if the destination already resolves to the source, skip copying and verify discovery instead.
2. For an independent existing destination, compare its whole tree to the source and privately back it up before resolving divergence. For a new target, preview rsync -an --itemize-changes /workspace/team-skills/release-notes/ "$HOME/.claude/skills/release-notes/". Confirm the preview includes only that reviewed tree.
3. Apply the same narrowly scoped copy with rsync -a /workspace/team-skills/release-notes/ "$HOME/.claude/skills/release-notes/". These commands omit deletion: obsolete destination files require a separate explicit deletion preview.
4. Compare all relative paths, file hashes, modes and symlink targets against the source, then use the configured Claude host’s documented discovery/reload mechanism. Report the verified installed tree and whether the active session actually discovered it. No other host folder is created by this example.
Skill liên quan
- task-coordination-strategies – Phân rã tác vụ phức tạp, thiết kế đồ thị phụ thuộc và điều phối công việc đa agent với…
- documentation-and-adrs – Ghi lại quyết định và tài liệu. Dùng khi đưa ra quyết định kiến trúc, thay đổi API công…
- team-communication-protocols – Giao thức nhắn tin có cấu trúc cho giao tiếp nhóm agent, gồm chọn loại tin nhắn, phê…
- docusign-automation – Tự động hóa tác vụ DocuSign qua Rube MCP (Composio): template, envelope, chữ ký, quản lý…
- doomsday-escrow – Xác minh tài sản thế chấp Proof-of-Escrow bằng mật mã (bằng chứng outpoint UTXO trên L1…
- dos-verify-done-claims – Trước khi chấp nhận khẳng định ‘xong / đã phát hành / đã sửa’ của agent, hãy đối chiếu…
- team-composition-patterns – Thiết kế đội agent tối ưu với heuristic xác định quy mô, cấu hình có sẵn và cách chọn…
- dropbox-automation – Tự động hóa quản lý tệp, chia sẻ, tìm kiếm, tải lên, tải xuống và thao tác thư mục trên…