Token导航 LogoToken导航

Agent Skill · 研究检索

bluesky-skill(蓝天技能)

bluesky-skill 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

bluesky-skill 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 17,769。

总安装 6,421周安装 273GitHub Stars 公开资料未说明下载 2,250核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.1

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:bluesky-skill(蓝天技能)
来源仓库:https://github.com/johannesseikowsky/bluesky-skill
安装命令:
openclaw skills install bluesky-skill
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install bluesky-skill

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

管理 Bluesky(bsky)账户的各项社交操作,如发帖、私信和通知。适用于内容发布、互动维护和平台数据分析等场景。支持搜索、关注、屏蔽等全功能接口调用,集成度高。需绑定有效账户凭证并遵守平台速率限制规则。bluesky-skill 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

建议定期检查权限范围,避免敏感操作引发安全警告。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install bluesky-skill;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

bluesky-skill 怎么安装?

在终端执行 openclaw skills install bluesky-skill。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

bluesky-skill 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

bluesky-skill 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

bluesky-skill 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

bluesky-skill 用的人多吗?

公开数据是累计安装 6,421 次,近一周安装 273 次,下载 2,250 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

bluesky-skill 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
bluesky-skill

蓝天技能 · 本页

6,421—接触敏感数据MIT-0有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
bluesky-skill
description
>-
allowed-tools
Bash Read Edit Write Glob Grep
metadata
openclaw
requires
env
[BLUESKY_HANDLE, BLUESKY_APP_PASSWORD]
bins
[python3]
primaryEnv
BLUESKY_HANDLE

Bluesky Account Management

Operate a Bluesky social media account via ./bsky <command> [args]. All output is JSON. Run from the project root.

Setup

Install dependencies:

pip install atproto python-dotenv

Requires .env at project root:

BLUESKY_HANDLE=your-handle.bsky.social
BLUESKY_APP_PASSWORD=xxxx-xxxx-xxxx-xxxx

App passwords: https://bsky.app/settings/app-passwords. For DMs, enable "Allow access to your direct messages".

Auth is automatic. A session cache is stored at ~/.bsky_session.json (contains an exported session token). Delete this file to force re-authentication or when revoking access.

JSON Output

Every command prints one JSON object to stdout. Parse with json.loads().

Response Schemas

Post object (returned by get, and inside arrays from timeline, search-posts, my-posts, thread):

{
  "uri": "at://did:plc:abc/app.bsky.feed.post/xyz",
  "cid": "bafyrei...",
  "author": {"handle": "alice.bsky.social", "did": "did:plc:abc", "display_name": "Alice", "avatar": "https://..."},
  "text": "Post content here",
  "created_at": "2026-03-14T10:00:00Z",
  "like_count": 5, "repost_count": 2, "reply_count": 1,
  "viewer": {"liked": "at://...like-uri or null", "reposted": "at://...repost-uri or null"},
  "embed": {"images": [{"alt": "...", "thumb": "...", "fullsize": "..."}], "external": {"uri": "...", "title": "...", "description": "..."}, "record": {"uri": "...", "text": "...", "author": {...}}} or null,
  "reply": {"parent_uri": "at://...", "root_uri": "at://..."} // only present on replies
}

Profile object (returned by profile, and inside arrays from search-users):

{
  "handle": "alice.bsky.social", "did": "did:plc:abc",
  "display_name": "Alice", "description": "Bio text", "avatar": "https://...",
  "followers_count": 100, "follows_count": 50, "posts_count": 200,
  "viewer": {"following": "at://...or null", "followed_by": "at://...or null", "blocking": null, "blocked_by": null, "muted": null}
}

Actor object (short profile, inside post authors, follower/following lists, notification authors):

{"handle": "alice.bsky.social", "did": "did:plc:abc", "display_name": "Alice", "avatar": "https://..."}

Notification object (inside array from notifications):

{
  "reason": "reply|like|repost|follow|mention|quote",
  "uri": "at://...", "cid": "bafyrei...", "is_read": false,
  "indexed_at": "2026-03-14T10:00:00Z",
  "author": {"handle": "...", "did": "...", "display_name": "...", "avatar": "..."},
  "record_text": "Their reply/post text (if applicable)",
  "reason_subject": "at://...the post they liked/reposted/replied-to (if applicable)",
  "subject_text": "Text of the subject post (if reason_subject exists)"
}

Conversation object (inside array from dm-list):

{
  "id": "convo-id", "unread_count": 2,
  "members": [{"handle": "...", "did": "...", "display_name": "...", "avatar": "..."}],
  "last_message": {"id": "msg-id", "text": "...", "sent_at": "...", "sender_did": "did:plc:..."} or null
}

DM message object (inside array from dm-read):

{"id": "msg-id", "text": "Message text", "sent_at": "2026-03-14T10:00:00Z", "sender_did": "did:plc:..."}

Feed object (inside array from feeds):

{
  "uri": "at://did:plc:abc/app.bsky.feed.generator/whats-hot",
  "cid": "bafyrei...", "did": "did:web:...",
  "creator": {"handle": "...", "did": "...", "display_name": "...", "avatar": "..."},
  "display_name": "What's Hot", "description": "Trending posts across Bluesky",
  "avatar": "https://...", "like_count": 12345, "indexed_at": "2026-03-14T10:00:00Z"
}

Command Response Keys

Each command returns these top-level keys:

CommandResponse keys
post{"uri", "cid"}
delete{"deleted"} (the URI)
like{"liked", "uri"} (post URI + like record URI)
unlike{"unliked"}
repost{"reposted", "uri"} (post URI + repost record URI)
unrepost{"unreposted"}
timeline{"feed": [{"post": <post>, "reason": {"type": "repost", "by": <actor>} or null}], "cursor"}
thread{"thread": <post with nested "replies": [...]>}
search-posts{"posts": [<post>], "cursor"}
search-users{"actors": [<profile>], "cursor"}
follow{"followed", "uri"}
unfollow{"unfollowed"}
followers{"followers": [<actor>], "cursor"}
following{"following": [<actor>], "cursor"}
mute{"muted"}
unmute{"unmuted"}
block{"blocked", "uri"}
unblock{"unblocked"}
profile<profile> (top-level, no wrapper)
get<post> (top-level, no wrapper)
my-posts{"posts": [<post>], "cursor"}
user-posts{"posts": [<post>], "cursor"}
likes{"likes": [{"actor": <actor>, "created_at": "..."}], "cursor"}
reposts{"reposted_by": [<actor>], "cursor"}
notifications{"notifications": [<notification>], "cursor"}
notif-read{"success": true}
dm-list{"conversations": [<convo>], "cursor"}
dm-read{"convo_id", "messages": [<dm>], "cursor"}
dm-send{"sent": true, "convo_id", "message_id"}
dm-mark-read{"success": true}
update-profile<profile> (top-level, no wrapper)
post-thread{"posts": [{"uri", "cid"}, ...]}
feeds{"feeds": [<feed>], "cursor"}

Important: Note that timeline wraps posts in feed[].post (with an optional reason), while search-posts and my-posts use posts[] directly.

Pagination

List commands support --cursor TOKEN. The response includes "cursor" (null = no more results).

  1. First call: omit --cursor
  2. Next page: pass the returned cursor as --cursor
  3. Stop when cursor is null

Errors

Errors return JSON with exit code 1:

{"error": "ERROR_TYPE", "message": "Human-readable description", "type": "SUBTYPE (for AUTH_ERROR)"}

Error types: NOT_FOUND, NOT_LIKED, NOT_REPOSTED, NOT_FOLLOWING, NOT_BLOCKING, FILE_NOT_FOUND, INVALID_ARGS, AUTH_ERROR.

Command Quick Reference

Posting

CommandDescription
post "text"Create a text post (max 300 graphemes)
post "text" --image photo.jpg --alt "description"Post with image (repeat --image/--alt for up to 4)
post "text" --reply-to <uri>Reply to a post
post "text" --quote <uri>Quote a post
post "text" --quote <uri> --image photo.jpg --alt "desc"Quote with image
post-thread "text1" "text2" "text3"Create a multi-post thread
delete <uri>Delete a post

Engagement

CommandDescription
like <uri>Like a post
unlike <uri>Unlike (pass the post URI, not the like URI)
repost <uri>Repost a post
unrepost <uri>Undo repost (pass the post URI)

Reading & Discovery

CommandDescription
timeline [--limit N] [--cursor TOKEN]Home timeline (default 20)
thread <uri> [--depth N]Post thread with replies (default depth 6)
search-posts "query" [--limit N] [--cursor TOKEN]Search posts
search-users "query" [--limit N] [--cursor TOKEN]Search users
feeds [--query "keyword"] [--limit N] [--cursor TOKEN]Browse suggested feed generators (note: --query filters client-side, so results may be fewer than --limit)

Social Graph

CommandDescription
follow <handle>Follow
unfollow <handle>Unfollow
followers <handle> [--limit N] [--cursor TOKEN]List followers
following <handle> [--limit N] [--cursor TOKEN]List following
mute <handle> / unmute <handle>Mute/unmute
block <handle> / unblock <handle>Block/unblock

Profile & Info

CommandDescription
profile [handle]View profile (own if omitted)
update-profile [--name "Name"] [--bio "Bio"] [--avatar image.jpg]Update your profile
my-posts [--limit N] [--cursor TOKEN]Own recent posts
user-posts <handle> [--limit N] [--cursor TOKEN]A user's recent posts
get <uri>Fetch a single post
likes <uri> [--limit N] [--cursor TOKEN]Who liked a post
reposts <uri> [--limit N] [--cursor TOKEN]Who reposted a post

Notifications

CommandDescription
notifications [--limit N] [--unread-only] [--filter TYPE] [--cursor TOKEN]List notifications (filter: like, repost, follow, mention, reply, quote)
notif-readMark all as read

Direct Messages

CommandDescription
dm-list [--limit N] [--cursor TOKEN]List conversations
dm-read --handle <handle> [--limit N] [--cursor TOKEN]Read messages with a user
dm-read --convo-id <id> [--limit N] [--cursor TOKEN]Read messages by convo ID
dm-send <handle> "text"Send a DM
dm-mark-read --convo-id <id>Mark convo as read
dm-mark-read --allMark all as read

Only text DMs are supported (no images).

Common Workflows

Check and respond to mentions

./bsky notifications --unread-only --filter mention
# Parse → for each notification, extract reason_subject (the post they mentioned you in)
./bsky get <reason_subject_uri>
# Read context, then reply:
./bsky post "Your reply" --reply-to <uri>
./bsky notif-read

Engage with timeline

./bsky timeline --limit 30
# Parse → extract feed[].post objects
# Like interesting posts:
./bsky like <uri>
# Reply to engage:
./bsky post "Your reply" --reply-to <uri>

Search and engage with a topic

./bsky search-posts "topic keywords" --limit 20
# Parse → like/repost/reply to relevant posts[]
./bsky like <uri>
./bsky repost <uri>

Join a conversation (read thread before replying)

./bsky thread <uri> --depth 6
# Parse → read thread.text and thread.replies[] to understand context
./bsky post "Informed reply" --reply-to <uri>

Grow the network

./bsky search-users "topic or niche" --limit 20
# Parse → review actors[] profiles for relevance
./bsky profile <handle>
# Check their posts before following:
./bsky user-posts <handle> --limit 10
# Avoid re-following — check viewer.following is null, then:
./bsky follow <handle>

Check engagement on own posts

./bsky my-posts --limit 10
# Parse → find posts with high reply_count
./bsky thread <uri>
# Respond to replies, like engaged followers
./bsky likes <uri>

Check and respond to DMs

./bsky dm-list
# Parse → find conversations[] with unread_count > 0
./bsky dm-read --convo-id <id>
# Parse → read messages[], reply:
./bsky dm-send <handle> "Your reply"
./bsky dm-mark-read --convo-id <id>

Post a thread

./bsky post-thread "First, let me explain the context..." "Second, here's the main point..." "Finally, the conclusion."
# Returns: {"posts": [{"uri": "...", "cid": "..."}, ...]}

Update your profile

./bsky update-profile --name "New Display Name" --bio "Updated bio text"
./bsky update-profile --avatar new-avatar.jpg

Discover feeds

./bsky feeds --limit 10
# Filter by keyword:
./bsky feeds --query "news"

Check before posting (avoid duplicates)

./bsky my-posts --limit 5
# Parse → check posts[].text for similar content
./bsky post "New post text"

Key Concepts

  • Handles: Always pass handles without the @ prefix — use user.bsky.social, not @user.bsky.social.
  • URIs: Every post has an AT Protocol URI (at://did:plc:abc/app.bsky.feed.post/xyz). Extract from the uri field in JSON. Used as arguments for like, reply, repost, thread, get, delete.
  • Rich text: @mentions, #hashtags, URLs in post text are auto-converted to links. Write naturally.
  • Character limit: 300 graphemes per post.
  • Unlike/unrepost: Pass the post URI, not the like/repost record URI. Auto-resolved internally.
  • Reply threading: --reply-to <uri> auto-resolves the thread root.

Auth Troubleshooting

Auth errors: {"error": "AUTH_ERROR", "type": "<TYPE>", "message": "..."} with exit code 1.

  1. SESSION_CORRUPT → rm ~/.bsky_session.json and retry
  2. MISSING_ENV → Ensure .env has BLUESKY_HANDLE and BLUESKY_APP_PASSWORD
  3. INVALID_CREDENTIALS → Handle: user.bsky.social, App password: xxxx-xxxx-xxxx-xxxx (19 chars)
  4. NETWORK → Retry up to 3 times with 10s delay
  5. ACCOUNT_SUSPENDED → Inform user, cannot fix programmatically

平台分布

OpenClaw

96.68%
按下载量换算2,175

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

可疑

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills