sandbaseai/sandbase-skills
88 installable open-source Agent Skills for research, social intelligence, marketing, and business workflows—compatible with Codex, Claude Code, Cursor, Gemini CLI, and DeepSeek Harness.
编辑导读Editor's Brief
## 核心特性
- 该仓库提供 88 个可安装的 Agent Skills,覆盖研究、社交情报、营销和业务流程。
- 每个技能都是一个指令文件,定义可重复的工作流、证据规则和输出格式。
- multi-source-search 技能可使用宿主提供的搜索和页面读取工具运行,无需 SandBase 账户。
- 离线校验器检查 evidence-ledger 一致性,并拒绝未知或重复来源、虚高置信度、未使用证据,以及存在已声明冲突的高置信度断言。
- 该仓库支持通过 npx skills、GitHub CLI、兼容 Copilot 的 Agent Plugins 1.0 包、DeepSeek Harness 捆绑包以及 Claude Code 市场插件进行安装。
## 使用场景
- 研究型 agent 可以跨 Web、学术和平台特定来源交叉核验断言,同时记录与来源关联的证据。
- 营销和竞争情报团队可以监控竞争对手、定价、活动、品牌提及和社交平台动态。
- 社交情报工作流可以收集并分析受支持平台上的帖子、趋势、创作者、评论和情绪。
- 实用工具任务可以解析文档、将页面转换为 Markdown、提取 YouTube 字幕、检查域名或验证邮箱。
## 技术细节
- 技能以附带参考资料的 Markdown 指令文件形式分发,可安装到实现 Agent Skills 规范的 agent 中。
- DeepSeek Harness 安装方式通过 DSH 的文件系统技能提供器挂载打包的 marketing/ 和 research/ 目录,同时保留默认技能根目录。
- 可选的 SandBase 提供器在环境中设置 SANDBASE_API_KEY 后,通过 MCP 风格工具访问。
- Copilot 插件按照 Agent Plugins 1.0 规范打包 multi-source-search 及其参考资料和离线 evidence-ledger 校验器。
## 注意事项
- 专用的社交、市场和数据提供器技能需要配置 SandBase,而 multi-source-search 可使用现有宿主搜索工具运行。
- 离线校验器仅检查内部一致性,不验证外部来源是否真实。
- 可选的 SandBase API 调用按用量计费,与 Apache-2.0 许可的技能分开计价。
## Core Features
- The repository provides 88 installable agent skills for research, social intelligence, marketing, and business workflows.
- Each skill is an instruction file that defines a repeatable workflow, evidence rules, and an output format.
- The multi-source-search skill can run with host-provided search and page-reading tools without a SandBase account.
- An offline validator checks evidence-ledger consistency and rejects unknown or duplicate sources, inflated confidence, unused evidence, and high-confidence claims with declared conflicts.
- The repository supports installation through npx skills, GitHub CLI, a Copilot-compatible Agent Plugins 1.0 package, a DeepSeek Harness bundle, and a Claude Code marketplace plugin.
## Use Cases
- Research agents can cross-check claims across web, academic, and platform-specific sources while recording source-linked evidence.
- Marketing and competitive teams can monitor competitors, pricing, campaigns, brand mentions, and social-platform activity.
- Social-intelligence workflows can collect and analyze posts, trends, creators, comments, and sentiment across supported platforms.
- Utility tasks can parse documents, convert pages to Markdown, extract YouTube transcripts, inspect domains, or validate emails.
## Technical Details
- Skills are distributed as Markdown instruction files with references and can be installed into agents that implement the Agent Skills specification.
- The DeepSeek Harness installation mounts packaged marketing/ and research/ directories through DSH's filesystem skill provider while preserving default skill roots.
- Optional SandBase providers are accessed through MCP-style tools after setting SANDBASE_API_KEY in the environment.
- The Copilot plugin packages multi-source-search with references and the offline evidence-ledger validator under the Agent Plugins 1.0 specification.
## Notes
- Specialized social, market, and data-provider skills require SandBase configuration, while multi-source-search can operate with existing host search tools.
- The offline validator checks internal consistency only and does not verify that an external source is true.
- Optional SandBase API calls are usage-based and priced separately from the Apache-2.0 licensed skills.
安装Install
dsh plugin --profile web add github:sandbaseai/sandbase-skills
把 sandbaseai/sandbase-skills 加入你的 DSH 配置(web profile)即可启用。
READMEREADME
SandBase Skills
Browse the flagship Skill: skills.sh · Agentic Awesome Skills · Agent Skill Exchange · askill · Awesome Skills · skills.re · SkillsCat · OpenAgentSkill · AgentSkill.sh
Have a repeatable workflow that is missing? Request a Skill with a public example and acceptance criteria.
English | 中文 | 日本語 | 한국어 | Español | Français | Deutsch | Português
88 installable Agent Skills for research, social intelligence, marketing, and business workflows. Install into any compatible agent (DeepSeek Harness, Claude Code, Codex, Cursor, Gemini CLI) and start working immediately. The flagship research Skill works with host-provided search tools; connect SandBase when you want broader provider coverage.
Start with multi-source-search: it runs with your agent's existing search tools,
ships a worked evidence-ledger example, and includes an offline validator. If it
improves a real workflow, star the repository
so other builders can discover it.
What are Skills?
A Skill is an instruction file that teaches an AI agent how to do one specific job. Each Skill defines a repeatable workflow, evidence rules, and output format. Portable Skills can use capabilities already provided by the host agent; specialized social, market, and data workflows can add SandBase providers when configured.
Quick Start
# Try it without installing: generate the complete Skill prompt
npx skills use sandbaseai/sandbase-skills@multi-source-search
# Or install it into Codex
npx skills add sandbaseai/sandbase-skills@multi-source-search --agent codex
# Use it with your agent's existing web/search tools
# "Fact-check this claim with independent sources and validate the evidence ledger"
multi-source-search needs no SandBase account when the host agent already provides
search and page-reading tools. For specialized social, market, and data-provider Skills,
set SANDBASE_API_KEY in your environment—never in a prompt or committed file.
GitHub CLI (official Agent Skills workflow)
GitHub CLI 2.90.0 or later can preview a Skill before installing it. Because this repository groups Skills by domain, pass the exact path shown in the catalog:
# Inspect the instructions and bundled files without installing anything
gh skill preview sandbaseai/sandbase-skills research/multi-source-search
# Install it for Codex at user scope
gh skill install sandbaseai/sandbase-skills research/multi-source-search \
--agent codex --scope user
Replace the final path with another linked Skill path from the catalog below. See
GitHub's gh skill documentation for
supported agents, version pinning, updates, and security guidance.
Portable Agent Plugin (Copilot CLI and compatible clients)
Install the curated, account-free research plugin directly from this repository:
copilot plugin install sandbaseai/sandbase-skills:agent-plugin
The plugin follows the vendor-neutral Agent Plugins 1.0 specification
and currently packages multi-source-search with its references and offline evidence-ledger
validator. Its generated Skill copy is checked against the canonical source in CI.
Add SandBase MCP tools to the same agent
Skills provide repeatable instructions; SandBase CLI adds six MCP tools for discovering and running 2,000+ models and APIs. Connect one of the 25 supported client targets, then use these Skills and MCP tools in the same agent:
npx -y https://github.com/sandbaseai/cli/releases/download/v0.1.17/sandbaseai-cli-0.1.17.tgz connect --client codex
Replace codex with another client ID from the
verified catalog.
DeepSeek Harness
Install all 88 Skills as a native DSH bundle:
dsh plugin --profile web add github:sandbaseai/sandbase-skills
dsh web
The bundle mounts the packaged marketing/ and research/ directories through
DSH's filesystem Skill provider while preserving the default Skill roots.
To copy only the flagship Skill into one project's
.dsh/skills/multi-source-search directory:
npx --yes github:sandbaseai/sandbase-skills add multi-source-search
The single-Skill installer and the bundle both run directly from the GitHub source, so no npm publication or SandBase account is required.
Claude Code marketplace
Install all 88 Skills as a native Claude Code plugin:
/plugin marketplace add sandbaseai/sandbase-skills
/plugin install sandbase-skills@sandbase-agent-skills
The marketplace manifest lists every Skill explicitly, so Claude Code can discover them on demand without copying directories by hand.
Verify research output offline
The multi-source-search Skill produces an
evidence ledger that can be checked before you trust or share its synthesis:
See the complete worked example, which cross-checks one claim against primary documentation from GitHub, GitLab, and Atlassian.
python3 research/multi-source-search/scripts/validate_report.py \
examples/verifiable-research-report.json
# VALID: 3 source(s), 1 claim(s), 2 provider(s)
The validator rejects unknown or duplicate sources, inflated confidence, unused evidence, and high-confidence claims that still have a declared conflict. It runs offline and checks internal consistency; it does not claim that a source is true.
For an end-to-end example that runs this research contract inside a governed, sandboxed runtime, read Build an Auditable Research Agent.
Try a Real Workflow
Install the matching Skill, then give your agent one of these tasks:
| Skill | Example task | What you get | Install |
|---|---|---|---|
twitter-intelligence |
“Compare sentiment and recurring complaints for Brand A and Brand B on X this month.” | Source-linked posts, trends, accounts, and a structured comparison | skills.sh |
multi-source-search |
“Verify the strongest evidence for and against this market claim.” | Cross-checked web and academic findings with disagreements called out | skills.sh |
competitor-monitor |
“Track three competitors’ pricing, launches, content, and social activity.” | A dated competitive-intelligence brief with observed changes | |
seo-content-brief |
“Build a writer-ready brief for this target keyword.” | Search intent, competing pages, required subtopics, and differentiation angles | skills.sh |
github-profile-research |
“Assess this engineering team’s open-source activity.” | Repository, language, contribution, star, and activity analysis | skills.sh |
youtube-research |
“Map the leading channels and audience questions in this niche.” | Video and channel discovery, transcript evidence, and comment themes | skills.sh |
Skill Catalog (88 Skills)
Social Intelligence (14 Skills)
Research and monitor conversations across every major social platform.
| Skill | Platform | Use it to |
|---|---|---|
twitter-intelligence |
Twitter/X | Search tweets, track trends, analyze users, monitor sentiment |
youtube-research |
YouTube | Search videos, analyze channels, extract transcripts, read comments |
instagram-research |
Analyze profiles, track hashtags, research content strategies | |
tiktok-research |
TikTok | Search videos, analyze creators, track hashtag challenges |
linkedin-research |
Research companies, professionals, job markets | |
reddit-research |
Search discussions, monitor communities, discover trends | |
xiaohongshu-research |
Xiaohongshu (RED) | Search notes, analyze creators, track consumer trends |
weibo-research |
Monitor hot searches, track trending topics, analyze sentiment | |
douyin-research |
Douyin | Search videos, analyze creators, track challenges |
wechat-channels-research |
WeChat Channels | Search videos, analyze live streams |
wechat-mp-research |
WeChat Official Accounts | Research articles, analyze accounts |
wechat-search |
Search across WeChat ecosystem | |
china-social-research |
Multi-platform | Cross-platform China social research |
community-research |
Reddit + Telegram | Online community analysis |
Search & Research (17 Skills)
Find, validate, and synthesize information from multiple sources.
| Skill | Use it to |
|---|---|
multi-source-search |
Cross-validate research with host search tools and optional Tavily, Exa, Scholar, and Cloudsway coverage |
tavily-deep-research |
Advanced web search with content extraction and site mapping |
exa-deep-search |
Semantic search and source extraction with Exa |
exa-similar-finder |
Find pages similar to any URL |
academic-research |
Search scholarly papers with AI-powered explanations |
academic-trend-research |
Track emerging research areas and breakthrough papers |
google-news-research |
Monitor news articles and media coverage |
web-scraper |
Scrape pages, crawl sites, extract structured data |
last30days-research |
Multi-platform research for the last 30 days |
topic-deep-dive |
Exhaustive multi-source topic research |
event-tracker |
Track events in real-time across platforms |
trend-spotter |
Spot emerging trends across platforms |
news-aggregator |
Aggregate news from multiple sources |
market-research |
Comprehensive market intelligence |
newsletter-research |
Discover industry newsletters and publications |
podcast-research |
Find podcasts and episodes by topic |
content-ideation |
Generate data-backed content ideas |
Business Intelligence (20 Skills)
Company research, sales intelligence, and competitive analysis.
| Skill | Use it to |
|---|---|
apollo-company-research |
Search companies, enrich profiles, track hiring |
akta-company-research |
Research companies via employee and product reviews |
company-enrichment |
Enrich company data from multiple sources |
google-maps-reviews |
Analyze local business reviews |
product-intelligence |
Cross-platform product market research |
product-review-extractor |
Extract and analyze product reviews at scale |
competitive-pricing |
Benchmark pricing against competitors |
pricing-page-analyzer |
Extract and analyze competitor pricing |
review-aggregator |
Aggregate reviews from multiple platforms |
startup-research |
Research startups with funding, team, and traction |
hiring-intelligence |
Analyze hiring patterns and talent competition |
talent-sourcing |
Source candidates by skills and expertise |
lead-research |
Build complete prospect profiles |
outreach-builder |
Build verified outreach lists |
email-outreach-prep |
Prepare personalized email outreach |
sales-intelligence |
Account intelligence for sales conversations |
partnership-research |
Qualify potential partners with data |
industry-landscape |
Map any industry's competitive landscape |
local-market-research |
Research local markets with reviews and social data |
data-enrichment |
Fill data gaps with verified intelligence |
Marketing & Content (15 Skills)
Brand monitoring, influencer marketing, and content strategy.
| Skill | Use it to |
|---|---|
brand-monitoring |
Track brand mentions across all platforms |
kol-discovery |
Find and evaluate influencers across platforms |
influencer-analytics |
Analyze influencer performance with engagement data |
social-listening |
Monitor conversations about any topic globally |
competitor-content-intelligence |
Find differentiated content opportunities |
competitor-monitor |
Monitor competitor launches, pricing, campaigns, and market signals |
thought-leadership-monitor |
Track industry voices and their content |
pr-media-monitor |
Track press mentions and media narrative |
content-performance |
Analyze content performance across platforms |
audience-research |
Understand target audiences from community data |
social-proof-research |
Find authentic testimonials and endorsements |
competitor-ad-research |
Research competitor advertising strategies |
hashtag-tracker |
Track hashtag performance across platforms |
crisis-monitor |
Detect and assess crises before they escalate |
reddit-customer-insights |
Discover customer language and pain points |
Marketing & SEO (5 Skills)
Search engine optimization, SERP analysis, and technical auditing.
| Skill | Use it to |
|---|---|
seo-keyword-insights |
Build evidence-backed keyword strategies with DataForSEO |
backlink-gap-analysis |
Find ethical backlink gaps against competitors |
serp-analysis |
Analyze live Google SERP results and features |
seo-content-brief |
Generate SERP-backed content briefs |
site-audit |
Audit website content, structure, and SEO health |
Tools & Utilities (17 Skills)
Practical tools for everyday agent tasks.
| Skill | Use it to |
|---|---|
email-validator |
Verify email deliverability and reputation |
domain-intelligence |
Research domains (WHOIS, DNS, SSL, security) |
domain-analyzer |
Complete domain analysis from DNS to SEO |
tech-stack-detector |
Identify what tech powers any website |
screenshot-capture |
Capture screenshots of any URL |
url-to-markdown |
Convert web pages to clean Markdown |
youtube-transcript |
Extract transcripts from YouTube videos |
document-parser |
Parse PDFs and documents to text |
content-translator |
Translate text between languages |
sentiment-analyzer |
Analyze sentiment in any text |
weather-lookup |
Check weather conditions worldwide |
flight-tracker |
Track flight status in real time |
currency-converter |
Convert currencies with live rates |
github-profile-research |
Research developer GitHub profiles |
npm-package-research |
Evaluate npm packages before installing |
cve-lookup |
Look up security vulnerabilities by CVE |
website-monitor |
Monitor websites for changes and health |
Install
# Install any Skill by name
npx skills add sandbaseai/sandbase-skills@<skill-name> --agent codex
# Global install (available across all projects)
npx skills add sandbaseai/sandbase-skills@<skill-name> --agent codex --global
# Find indexed SandBase Skills by topic
npx skills find "research" --owner sandbaseai
# Browse every Skill in the repository
npx skills add sandbaseai/sandbase-skills --list
Supported Agents
Skills work with any agent that implements the Agent Skills specification:
- Claude Code —
~/.claude/skills/ - OpenAI Codex —
~/.codex/skills/ - Cursor —
~/.cursor/skills/ - Gemini CLI —
~/.gemini/skills/ - OpenClaw, Hermes, Amp, Devin — via
npx skills add
How It Works
User Question → Agent reads SKILL.md → Uses host tools and/or SandBase → Validates evidence → Answer
- You ask a question or give a task
- Your agent reads the installed Skill's instructions
- The Skill starts with compatible search or browser tools already available to the host
- When configured, SandBase adds provider-specific capabilities via
sandbase_describe_tool→sandbase_call_tool - The agent executes the workflow, validates structured evidence, and delivers the result
SandBase Ecosystem
- SandBase CLI — connect Codex, Claude Code, Cursor, Gemini CLI, and other MCP clients to 2,000+ AI models with one onboarding command.
- SandBase Harness — run persistent agent sessions with sandboxed tools, resumable streams, artifacts, cancellation, audit, and replay in your own infrastructure.
- DSH Plugin Store — discover runtime-verified DeepSeek Harness plugins, install them into a local Web profile, and inspect active loader state from Settings.
Pricing
Skills themselves are free and open source (Apache-2.0). multi-source-search
can use compatible host-provided tools without a SandBase account. Optional SandBase
API calls are usage-based — typically $0.001–$0.01 per call. A typical research task
using those calls costs $0.05–$0.20.
See sandbase.ai/pricing for current rates.
Repository Layout
research/<skill>/SKILL.md Agent instruction file
research/<skill>/references/ API maps and workflow guidance
marketing/<skill>/ Original marketing skills
catalog/skills/ Web display metadata
integrations/sandbase-registry/ Platform registry manifests
scripts/skillpack.py Validation helper
Contributing
See CONTRIBUTING.md before adding or changing a Skill.
# Validate locally
python3 scripts/skillpack.py validate
python3 -m unittest discover -s tests -v
License
Apache-2.0. See LICENSE.
Built for SandBase — One API key. Every data source. 88 installable agent skills.
amruthpillai/reactive-resume
tt-a1i/archify
strukto-ai/mirage
liustack/modlens
omdsh-dev/DSH-better-sidebar
dsh-market/dsh-market
agentrq/agentrq