Capture raw content (metadata, transcript, frames) from YouTube videos and Instagram reels, analyze it in the current Claude session, download YouTube videos (full or trimmed), and extract multiple clips from long-form content (podcasts, courses) via a conversational brainstorm. Skills: view-youtube-video, view-instagram-reel, extract-content, clip-youtube-video, clip-video-parts.
0 on GitHubby Paulo Soaresv0.1.0MITupdated 11 days agocontent
/
Plugin psoares-content-extract from marketplace psoares-claude-plugins in ps-dev-team/psoares-claude-plugins. Commands read from the repo's manifest — 11 days ago.
What it installs
Skills (5)
clip-video-parts
Analyze a long video or podcast (local file or YouTube URL), brainstorm with the user which parts are worth extracting, and cut multiple clips in one pass. Use when the user has a 30-min+ recording and wants several standalone clips — not just one range. Trigger phrases: "saca partes deste vídeo", "
clip-youtube-video
Download a YouTube video — either the whole thing or a specific time range — to a local file. Use when the user wants the actual video file, not an analysis. Trigger phrases: "descarrega este vídeo", "saca este youtube", "quero o vídeo", "download this youtube", "clip from X to Y", "saca de 1:20 a 2
extract-content
Analyze a captured social-media video (YouTube or Instagram) and extract insights — summaries, action items, deep dives, repurposing ideas, or full transcript. Use when the user asks to analyze, summarize, break down, or extract value from a video that has already been ingested by the view skills, O
view-instagram-reel
Capture raw content from an Instagram Reel (or regular post) so later skills can analyze, repurpose, or summarize it. Use when the user shares an Instagram URL and wants to ingest it — trigger phrases like "view this reel", "captura este reel", "processa este post do instagram", "vê este reel", "ing
view-youtube-video
Capture raw content from a YouTube video so later skills (extraction, analysis, repurposing) can work on it. Use when the user shares a YouTube URL and wants to ingest it — trigger phrases like "view this video", "ingere este vídeo", "vê este vídeo do youtube", "vê este short do youtube", "captura e
Screen demos and B-roll for creators, from a script or plain instructions: screen (real headed Chrome, one or two windows, beats timed to a voice or to seconds, the pointer moving like a hand), assemble (large Screen Studio-style pointer, spring zooms, plain browser frame, fades, 4K 60 fps, normalized loudness), audio (loudness, peaks, pauses and speech rate; normalize to -16 or -14 LUFS; pause cap), voice (optional: generate with ElevenLabs, or bring your own audio or video and get line and word timings; shorter pauses), broll (optional, paid: people B-roll with Higgsfield). macOS, Linux and Windows.
Writing for videos and content. Skills: hooks for the opening (six formats, the spoken, text and visual hook lined up, the lock-in zone, four mistakes), storytelling for the structure that keeps people watching (stakes, a big question, head fakes, re-hooks), and video-script for scripts recorded one take per line.
Development workflow tooling. Skills: create-github-issue for filing grounded, well-scoped GitHub issues from the codebase, and journal-session for recording what happened in a Claude session as a narrative project journal.
Voices with ElevenLabs that sound like the speaker: eleven_v4 clones through text-to-dialogue, script tags and punctuation for pacing, pronunciation fixes that never touch the script, per-language pause caps, picking website takes from history, word timings and misread checks with Scribe. Script in, voice and line timings out.
The Markdown Teleprompter app's script format. Skill: teleprompter-script, which lays out a script the way the app reads it: directives, chapters, presenter tabs, presenter notes, voice cues, and bullet points to talk from instead of reading word for word.
The diff pane as a plugin: /diff opens the session's uncommitted changes beside the transcript, file by file with their hunks, refreshed as Claude edits and runs commands; the first edit opens it where the terminal is wide enough, and a file's ask button rides its hunks on the next prompt.
Research any topic across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and 5+ more sources. AI agent scores by upvotes, likes, and real money - not editors.
Web scraping, Google search, structured data extraction, and MCP server integration powered by Bright Data. Includes 7 skills: scrape any webpage as markdown (with bot detection/CAPTCHA bypass), search Google with structured JSON results, extract data from 40+ websites (Amazon, LinkedIn, Instagram, TikTok, YouTube, and more), orchestrate Bright Data's 60+ MCP tools, built-in best practices for Web Unlocker, SERP API, Web Scraper API, and Browser API, Python SDK best practices for the brightda...
Local session memory for coding agents: indexes existing session transcripts on disk across 43 coding agents into a local BM25 store with MCP recall, hooks, skill, and /deja command
Plan, build, and debug Zoom integrations across REST APIs, Meeting SDK, Video SDK, webhooks, bots, and MCP workflows. Search meetings, retrieve recordings, access transcripts, and design AI-powered Zoom experiences.
44 skills · 3 MCP servers
anthropics/knowledge-work-plugins 35,601
Maintain psoares-content-extract?
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.