Token导航 LogoToken导航

Agent Skill · 研究检索

socialcrawl(socialcrawl 搜索)

socialcrawl 用于处理浏览器自动化、网页检查和页面信息提取,适合在 Codex、Claude、Cursor、Gemini CLI 中需要让 Agent 打开页面、读取网页或验证前端流程时使用。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。

socialcrawl 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描有未通过项,按安装量在研究检索类的 73,675 个 Skill 里排第 44,445。

总安装 915周安装 37GitHub Stars 公开资料未说明下载 287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
unknown

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:socialcrawl(socialcrawl 搜索)
来源仓库:https://github.com/socialcrawl/skills
仓库路径:skills/socialcrawl
安装命令:
npx skills add https://github.com/socialcrawl/skills --skill socialcrawl
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 npx skills 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。

skills.shnpx skills
npx skills add https://github.com/socialcrawl/skills --skill socialcrawl

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。

socialcrawl 用于处理浏览器自动化、网页检查和页面信息提取,适合在 Codex、Claude、Cursor、Gemini CLI 中需要让 Agent 打开页面、读取网页或验证前端流程时使用。它支持网页内容抓取、DOM 结构解析和前端交互模拟,适用于需要快速获取页面数据或测试 UI 行为的场景。

通过 GitHub 仓库安装,使用 npx skills add 命令添加指定技能,具体用法可参考原始 README 文档。安装前需确认权限范围、维护状态,并注意可能触发的联网、命令执行或文件读写操作。建议结合来源仓库进一步核验功能细节,确保符合实际使用环境和安全要求。

来源仓库内路径:skills/socialcrawl

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 Codex、Claude、Cursor、Gemini CLI;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 npx skills add https://github.com/socialcrawl/skills --skill socialcrawl;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

socialcrawl 怎么安装?

在终端执行 npx skills add https://github.com/socialcrawl/skills --skill socialcrawl。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

socialcrawl 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。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。

socialcrawl 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。

socialcrawl 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

socialcrawl 用的人多吗?

公开数据是累计安装 915 次,近一周安装 37 次,下载 287 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。

socialcrawl 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

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。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 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 Codex,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。
  • 需要明确授权才能用到公司项目里的场景。
  • 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
socialcrawl

socialcrawl 搜索 · 本页

915—接触敏感数据未明确未通过

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

SocialCrawl API

Unified social media data API. One API key, one response format, 21 platforms, 108 endpoints. Author and Post responses are normalized through platform field maps and augmented with deterministic computed fields (engagement_rate, language, content_category, estimated_reach) under data.computed. List archetypes (PostList, CommentList, SearchResult, Audience, Analytics) pass through as {items, next_cursor?, total?} without computed fields. Add ?format=raw to bypass the transform pipeline entirely.

API Key

Resolve the API key before making any call, checking these sources in order:

  1. Env var: echo "$SOCIALCRAWL_API_KEY" — if set and starts with sc_ (and is not a placeholder like sc_your_api_key_here), use it.
  2. Config file: cat ~/.config/socialcrawl/api_key 2>/dev/null — if the file exists and contains a key starting with sc_, use it.
  3. Ask the user: If neither source has a valid key:

- Tell the user: "I need your SocialCrawl API key to continue. You can find it at https://socialcrawl.dev/dashboard — every account starts with 100 free credits." - Ask them to paste their key. - After receiving the key, auto-save it so they never need to paste it again: mkdir -p ~/.config/socialcrawl && echo "sc_xxxxx" > ~/.config/socialcrawl/api_key - Tell the user: "I've saved your key to ~/.config/socialcrawl/api_key so it will be available in future sessions."

For all subsequent API calls in the session, use the resolved key directly in the curl command (do not rely on the env var being set).

First Use

On the first interaction with this skill in a session:

  1. Briefly introduce: "SocialCrawl provides a single API for 21 social media platforms (105 endpoints). Let me verify your API key."
  2. Resolve the API key using the steps above. If the key is missing or a placeholder, stop here and ask for it before proceeding.
  3. Tell the user you'll make a test call that costs 1 credit, then run: curl -s -H "x-api-key: KEY" "https://www.socialcrawl.dev/v1/tiktok/profile?handle=tiktok" (Replace KEY with the resolved key value.)
  4. If successful, confirm the key works and show credits_remaining. Then respond to whatever the user actually asked.
  5. If it fails, report the error and help troubleshoot (see Error Handling below).

Platforms

Workflow

Determine what the user wants, then follow the matching workflow:

User wants data:

  1. Identify the platform and resource from their request
  2. Read the platform's reference file from the table above
  3. Resolve API key
  4. Construct and execute the curl command
  5. Return raw JSON response
  6. Note credits_used and credits_remaining from the response

User wants code:

  1. Identify platform, resource, and target language
  2. Read the platform's reference file
  3. Generate a working code snippet using $SOCIALCRAWL_API_KEY env var for the key
  4. Present the code without executing

User asks about capabilities:

  1. Answer from the platform table above
  2. If they need details about auth, response format, errors, or credits, read references/api-overview.md

User asks about credits/balance:

  1. Resolve API key
  2. Run: curl -s -H "x-api-key: $SOCIALCRAWL_API_KEY" "https://www.socialcrawl.dev/v1/credits/balance"
  3. Return data.balance from the envelope (the call costs 0 credits)

Ambiguous platform: If the user says "get profile for @nike" without specifying a platform, ask which platform they mean.

Multi-platform requests: Load each platform's reference file and make sequential calls.

Making API Calls

Base URL: https://www.socialcrawl.dev

All endpoints are GET requests:

curl -s -H "x-api-key: $SOCIALCRAWL_API_KEY" \
  "https://www.socialcrawl.dev/v1/{platform}/{resource}?{param}={value}"

URL-encode parameter values that contain spaces or special characters.

Parameter rules

  • Required params — each platform reference file lists the required query parameters for every endpoint. A request missing any of them returns 400 INVALID_REQUEST with the message Missing required parameter(s):...; no credit is deducted.
  • oneOf groups — some endpoints (notably on YouTube, Facebook, TikTok, Instagram, Reddit, Google, Truth Social) accept mutually-substitutable identifiers such as channelId / handle / url. Pass at least one member of the group; the platform reference file lists the valid choices.
  • Optional params — any non-required param listed in the reference file may be forwarded. Anything unrecognized is rejected upstream.
  • format=raw — disables field maps and computed fields so you receive the original ScrapeCreators JSON.

Credit Tiers

TierCostTypical endpoints
standard1 creditProfiles, posts, search, comments
advanced5 creditsAudience demographics, ad libraries, trending
premium10 creditsVideo transcripts, age/gender detection

Before executing an advanced or premium call, mention the credit cost to the user. After every call, report credits_used and credits_remaining from the response.

Free calls (0 credits deducted):

  • Cache hits — cached: true + X-Cache: HIT. Same data, no charge.
  • Idempotent replays — X-Idempotent-Replay: true. See "Idempotent Retries" below.
  • Empty-upstream 404s — when upstream returns a 200 with an empty body (nonexistent profile/post), the credit is auto-refunded and you get RESOURCE_NOT_FOUND.
  • GET /v1/credits/balance — meta endpoint.
  • Pre-flight rejections — 400/401/402/405/409/422/404-endpoint-not-found never deduct.

Idempotent Retries

Any /v1/* call can be made safe to retry by sending an Idempotency-Key header (UUIDv4 or any opaque 16+ char string). Replays of the same key + same params return the original body verbatim with X-Idempotent-Replay: true and 0 credits deducted. Keys last 24h. Reusing the key with different params returns 422; a key owned by a different account returns 409.

curl -s -H "x-api-key: $SOCIALCRAWL_API_KEY" \
  -H "Idempotency-Key: 7a5e1b4c-2d8f-4a3b-9c1e-6e8b4d2a1f3c" \
  "https://www.socialcrawl.dev/v1/tiktok/profile?handle=tiktok"

Response Headers

HeaderValue
X-Request-Idreq-XXXXX
X-Credits-UsedCredits charged (0 on cache hit, replay, or refund)
X-Credits-RemainingBalance after this call
X-CacheHIT or MISS
X-Idempotent-Replay"true" on replays (absent otherwise)
Retry-After"30" — only on 503 circuit-breaker responses
Allow"GET" — only on 405 METHOD_NOT_ALLOWED

Warnings Channel

Successful responses may include an optional data._warnings: string[] with advisory notices from the transform pipeline (e.g. a clamped engagement_rate > 1.0, or an unresolved field-map path). Treat as observability-only — do not gate logic on it. Empty arrays are omitted.

Error Handling

CodeStatusAction
MISSING_API_KEY401Ask user for their API key
INVALID_API_KEY401"Your API key appears invalid. Check your SocialCrawl dashboard."
INSUFFICIENT_CREDITS402"You're out of credits. Top up at socialcrawl.dev/dashboard/billing"
INVALID_REQUEST400Missing required param OR malformed handle/URL (format validator). Check required params + formats in the platform reference.
METHOD_NOT_ALLOWED405All /v1/* endpoints are GET-only. Response includes Allow: GET.
ENDPOINT_NOT_FOUND404"That endpoint doesn't exist. Check the platform table above."
RESOURCE_NOT_FOUND404"That profile/post wasn't found on the platform." Credits were refunded if the upstream returned an empty body.
IDEMPOTENCY_KEY_CONFLICT409"That Idempotency-Key is in use by a different account. Pick a new key."
IDEMPOTENCY_KEY_PAYLOAD_MISMATCH422"You reused an Idempotency-Key with different params. Use a fresh key."
CONCURRENCY_LIMIT429"Too many concurrent requests (limit 50/key). Wait a moment and retry."
UPSTREAM_ERROR502"Platform temporarily unavailable. Credits were refunded."
SERVICE_UNAVAILABLE503"Platform circuit breaker is open. Try again in 30s. Credits refunded."
INTERNAL_ERROR500"Unexpected error. Credits were refunded."

References

  • references/api-overview.md — Read when user asks about authentication, response format, error details, credit system, or the ?format=raw parameter
  • references/{platform}.md — Read the specific platform file when user asks about or wants to call that platform's endpoints

平台分布

Codex

33%
按下载量换算95

Claude

31.09%
按下载量换算89

Cursor

20.08%
按下载量换算58

Gemini CLI

8.97%
按下载量换算26

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

未通过

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