mcp-dir avatar

Social → Context

mcp-dir/socialdl-mcp

Turns social media links into structured text context for AI agents. Drop in a TikTok, Instagram reel/post/carousel, or YouTube URL and get back metadata, audio/video transcription, and OCR plus vision summaries of images. Exposes three tools: socialdl_extract for full content conversion, socialdl_metadata for lightweight caption and author lookup without transcription, and socialdl_download for direct CDN URLs. YouTube pulls existing captions when available. First three calls are free, then usage-based pricing kicks in. Reach for this when you need Claude or your agent to understand what's inside a video or post without watching it yourself. Hosted service over streamable HTTP, so works across Claude Desktop, Cursor, VS Code, and any MCP-over-HTTP client.

tools
8
transport
HTTP
last updated
Jun 10, 2026

Install

From registry metadata
$ npx add-mcp https://api.mcp.ai/p_socialdl --name socialdl-mcp

Run in your terminal. add-mcp asks which agents to add it to (Claude Code, Codex, Cursor and more). Add -g to install it for every project.

Review the command, arguments and environment values before installing. MCP servers run with your local permissions.

Other ways

Claude Code

$ claude mcp add --transport http socialdl-mcp https://api.mcp.ai/p_socialdl

Run in your terminal. Add --scope user to make it available in every project.

Tools

Verified live · 8 tools

Verified live against the running server on Nov 6, 2026.

socialdl_extractDownload a social-media link (TikTok, Instagram reel/post/carousel, YouTube, or any direct .mp4) and return it as usable text context: metadata (author, caption, duration), the audio/video transcript, and a vision description (OCR + summary) of each image in a carousel/post. Y...4 params

Download a social-media link (TikTok, Instagram reel/post/carousel, YouTube, or any direct .mp4) and return it as usable text context: metadata (author, caption, duration), the audio/video transcript, and a vision description (OCR + summary) of each image in a carousel/post. Y...

Parameters* required
describe_imagesboolean
languagestring
transcribeboolean
url*string
socialdl_metadataLightweight lookup: resolve a social-media link to its metadata (author, caption, duration, counts) and media URLs WITHOUT transcribing or running vision. Cheaper than socialdl_extract — use when the caption/context alone is enough.1 params

Lightweight lookup: resolve a social-media link to its metadata (author, caption, duration, counts) and media URLs WITHOUT transcribing or running vision. Cheaper than socialdl_extract — use when the caption/context alone is enough.

Parameters* required
url*string
socialdl_downloadResolve a TikTok / Instagram / YouTube link (or direct .mp4/image URL) to its DIRECT download URLs from the source CDN — every video quality, audio track and image, with quality labels. Use when the user wants to DOWNLOAD/save the media instead of transcribing it. No transcrip...1 params

Resolve a TikTok / Instagram / YouTube link (or direct .mp4/image URL) to its DIRECT download URLs from the source CDN — every video quality, audio track and image, with quality labels. Use when the user wants to DOWNLOAD/save the media instead of transcribing it. No transcrip...

Parameters* required
url*string
show_versionShow the current MCP platform and adapter versions.

Show the current MCP platform and adapter versions.

No parameters. Call it with no arguments.

report_bugReport a bug, missing feature, or send feedback. Include the conversation array with recent messages for reproduction.3 params

Report a bug, missing feature, or send feedback. Include the conversation array with recent messages for reproduction.

Parameters* required
contextstring
default:
conversationstring
default: []
message*string
connectReturns connection status and URLs. When all providers are connected, returns authenticated:true and empty pending[]. When credentials are missing, returns connect_url for the toolkit and per-install URLs.

Returns connection status and URLs. When all providers are connected, returns authenticated:true and empty pending[]. When credentials are missing, returns connect_url for the toolkit and per-install URLs.

No parameters. Call it with no arguments.

toolkit_infoReturns the current toolkit state: installed MCPs, their connection status, the accounts connected to each one, and how many catalog tools each exposes.

Returns the current toolkit state: installed MCPs, their connection status, the accounts connected to each one, and how many catalog tools each exposes.

No parameters. Call it with no arguments.

marketplaceThe official mcp.ai marketplace — the in-platform catalog of every MCP/tool, AND the way to run them. Covers capability requests like "find an MCP that does X", "consulta um CPF", "is there a tool for Y". Core flow: action=search discovers MCPs by intent → describe returns one...23 params

The official mcp.ai marketplace — the in-platform catalog of every MCP/tool, AND the way to run them. Covers capability requests like "find an MCP that does X", "consulta um CPF", "is there a tool for Y". Core flow: action=search discovers MCPs by intent → describe returns one...

Parameters* required
actionstring
one of search · describe · install · uninstall · subscribe · canceldefault: search
argumentsstring
default: {}
cancel_commentstring
default:
cancel_reasonstring
one of too_expensive · missing_features · switched_service · unused · customer_service · too_complex
conversationstring
default: []
immediateboolean
default: false
limitnumber
default: 10
mcp_idstring
default:
messagestring
default:
prompt_bodystring
default:
prompt_categorystring
default:
prompt_descriptionstring
default:
prompt_slugstring
default:
prompt_targetsarray
prompt_titlestring
default:
prompt_toolstring
default:
prompt_varsstring
default: {}
querystring
default:
report_contextstring
default:
request_detailsstring
default:

+3 more parameters. See the server's own documentation for the full schema.

Sponsors

Readme

Social → Context

Transforme vídeos e posts de redes sociais em texto para Claude, Cursor e agentes de IA

Cole um link de rede social (TikTok, Instagram reel/post/carrossel, YouTube ou um .mp4) e receba o conteúdo como texto: metadados, transcrição do áudio/vídeo e descrição (OCR e resumo) das imagens. Dá ao Claude o contexto do que está dentro do vídeo ou post.

  • 🎬 Vídeo e post viram texto: transcrição do áudio + OCR e resumo das imagens
  • 🔗 TikTok, Instagram (reel/post/carrossel), YouTube ou qualquer .mp4
  • 🧠 Contexto pro seu agente: o Claude passa a entender o que está dentro do link
  • 🪶 Metadados leves (autor, legenda, duração) sem baixar o vídeo inteiro
  • 🔒 Somente leitura · 🎁 3 chamadas grátis
  • 💬 Funciona com qualquer cliente MCP: Claude Desktop, Cursor, VS Code, Cline, Continue

English version · Documentação completa · Skill pra agentes


Instalar em 1 clique

Claude (Web e Desktop)

A Anthropic unificou a instalação de MCPs em claude.ai/customize/connectors. O mesmo link serve pra Claude Web e Claude Desktop (basta estar logado):


View the full README on GitHub

Details

Repository
github.com/mcp-dir/socialdl-mcp
Category
Media & Entertainment
Transport
HTTP
Registry
MCP registry, active
Tools verified
Nov 6, 2026

More from mcp-dir

Related Media & Entertainment MCP servers

View all
1micro-jay avatarTwelvelabs Mcp Servermicro-jay/twelvelabs-mcp-serverMCP server for ElevenLabs Conversational AI (not TwelveLabs video) — agents, KB, voices0stars
2mkirovbg avatarScribefymkirovbg/scribefy-mcpSearch YouTube, get video metadata, and extract timestamped transcripts for AI workflows.0stars
3mobileshop9991-star avatarClipwisemobileshop9991-star/clipwise-mcpMCP server for Clipwise - AI tools for TikTok, Reels, and Shorts video creators.0stars
4okdataapi avatarokdataokdataapi/okdataSocial data for AI agents: Douyin, Xiaohongshu, WeChat, TikTok, X, YouTube, more. Pay per call.0stars
5pitchghost avatarPitchGhostpitchghost/mcpAI-powered social listening across Twitter, LinkedIn, Reddit, Facebook, and more.0stars
6rocnubie avatarEarScribe MCProcnubie/earscribe-mcpEarScribe is a browser-based transcription tool that converts audio to text on-device using OpenAI0stars