damionrashford

Trim, cut, split, segment, and concatenate media with ffmpeg (stream copy when possible, re-encode across cut boundaries). Use when the user asks to trim a video, cut a clip, extract a segment by timestamps, remove a section, split into parts, join/merge videos, concatenate files, or build a segmented HLS-style playlist.

Install and configure OBS Studio programmatically: install via brew cask / winget / Flatpak / apt, author profiles (basic.ini, streamEncoder.json, recordEncoder.json), author scene collections (scenes JSON), manage global.ini, set defaults for encoder / output / audio / hotkeys, cross-platform config paths. Use when the user asks to install OBS, set up an OBS profile, create a scene collection from code, configure OBS defaults without the GUI, edit basic.ini, manage multiple OBS profiles, or script a fresh OBS install with known-good settings.

Deep media inspection + automated QC — ffprobe stream details, MediaInfo diagnostics, VMAF/PSNR/SSIM quality metrics, PySceneDetect scene cuts, crop/silence/black-frame/interlacing detection, ffplay scope debugging, NAL/SEI bitstream forensics, metadata audits, loudness compliance against Spotify/Apple/ATSC/EBU specs, and automated CI QC gates. Use when the user says "QC this file", "run VMAF", "compare encoders", "detect scene cuts", "check loudness compliance", "validate delivery spec", "automated QC pipeline", or any deep inspection / quality gating.

Routes media production requests to the right Media OS specialist subagent. ALWAYS use this skill when the user expresses ANY media production intent — "go live", "start streaming", "OBS broadcast", "wire up live rig", "NDI to stream", "PTZ camera setup", "DeckLink capture stream", "HLS deliver", "DASH package", "encode for streaming", "CDN upload", "make a HLS manifest", "multi-bitrate ladder", "CMAF package", "LL-HLS", "low-latency stream", "Widevine package", "DRM package", "package for streaming", "broadcast deliver", "MXF master", "IMF package", "ProRes master", "DPP deliver", "AS-11 deliver", "Netflix deliver", "broadcast spec deliver", "deliver for air", "create IMF", "Premiere to Resolve", "round-trip", "OTIO export", "editorial conform", "XML round-trip", "FCPXML", "EDL export", "AAF export", "convert timeline", "Avid to Premiere", "Resolve to Premiere", "upscale this", "interpolate frames", "denoise AI", "remove background", "rotoscope", "matte", "depth estimate", "AI upscale", "RIFE", "Real-ESRGAN"

Ingest from every source — web (yt-dlp), screen / webcam / mic capture, SDI (DeckLink), DSLR tether (gphoto2), NDI network sources, RTSP / IP cameras via MediaMTX, PTZ — then verify integrity (SHA-256 + full-decode pass), preserve metadata (EXIF / XMP / IPTC), normalize to MKV or FFV1 / J2K / ProRes archival containers, and push to cold-storage cloud (Glacier / B2 / Archive.org). Use when the user says "download YouTube playlist", "capture SDI for 24 hours", "archive IP cameras", "preserve VHS rips", "tether DSLR timelapse", "cold-storage upload", or any ingest-to-archive workflow.

Restore, upscale, and enhance existing footage using 2026 open-source AI models — Real-ESRGAN/SwinIR/HAT super-resolution, RIFE/FILM interpolation, DeepFilterNet/RNNoise audio denoise, rembg/BiRefNet/RVM matting, Depth-Anything v2 depth — with strict OSI-open commercial-safe license filter. Use when the user says "upscale old footage", "remaster", "enhance quality", "30 to 60fps", "AI denoise", "restore VHS", "remove background from video", or anything about AI-driven footage restoration.

Generate media from scratch with 2026 open-source AI — TTS voiceover (Kokoro / OpenVoice / Piper), image gen (FLUX-schnell / Kolors / Sana / ComfyUI), video gen (LTX-Video / CogVideoX / Mochi / Wan), music (Riffusion / YuE), lipsync talking heads (LivePortrait / LatentSync), OCR (PaddleOCR / Tesseract 5 / TrOCR), zero-shot tagging (CLIP / SigLIP / BLIP-2 / LLaVA). Strict commercial-safe license filter. Use when the user says "generate a video", "TTS voiceover", "AI explainer video", "clone my voice", "generate music", "AI image", "digital human", or anything about from-scratch AI media.

Full-stack audio routing across macOS / Linux / Windows, MIDI + OSC control surfaces, DAW integration over JACK, DSP (EQ / compression / loudness / spatial / binaural HRTF), AI denoise (DeepFilterNet / RNNoise / Resemble Enhance), stem separation (Demucs), TTS, Whisper transcription, and EBU R128 / Spotify / Apple / ATSC loudness certification. Use when the user says "set up DAW", "route audio through OBS", "MIDI control surface", "binaural ASMR", "stem separation", "loudness to −16 LUFS", "JACK transport", or anything audio-stack-related.

Comprehensive Claude Code knowledge base — plugins, hooks, skills, agents, MCP, channels, headless mode, permissions, settings, and all extensibility features. Use when building, configuring, debugging, or extending Claude Code.

Use when building a skill, creating a SKILL.md, packaging a workflow, making a slash command, or asked "how do I make a skill". Scaffolds the folder, generates SKILL.md from a template, validates against spec. Produces a complete ready-to-deploy skill folder: scripts, references, assets. Also use to review or improve an existing skill.

Statistical analysis, hypothesis testing, A/B testing, cohort analysis, segmentation, trend detection, business metrics, pre-delivery validation, and data visualization. Use when the user asks to "analyze this data", "run a statistical test", "compare groups", "find trends", "do A/B test analysis", "segment customers", "calculate KPIs", "validate this analysis", "check my work", "sanity check", "review my numbers", "make a chart", "create a dashboard", "plot the data", "visualize results", or mentions hypothesis testing, cohort analysis, business analytics, data validation, bar charts, line charts, heatmaps, scatter plots, or data storytelling.

Autonomous time-budget experiment loop. Modify a training script, train for a fixed wall-clock budget, evaluate, record, repeat. Inspired by karpathy/autoresearch. Use for overnight architecture search, systematic hyperparameter sweeps, or any iterative model improvement workflow.

Context engineering for building production LLM applications: context window management, degradation patterns, optimization strategies, memory system selection, multi-agent architecture, filesystem context patterns, and tool design principles. Use when building LLM apps, RAG pipelines, AI agents, multi-agent systems, or when designing memory, tool APIs, or context strategies for any language model application.

Explore, clean, and engineer datasets end-to-end: statistical profiling, distribution checks, missing value analysis, duplicate detection, outlier removal, type fixing, encoding, create features, encode categories, transform columns, add rolling windows, build interaction terms, and feature engineering. Supports pandas, polars, and PySpark. Use when the user wants to explore data, profile columns, understand a dataset, clean data, handle missing values, remove duplicates, fix data types, preprocess a dataset before modeling, create features, encode categories, transform columns, add rolling windows, build interaction terms, or do feature engineering.

Detect data drift, concept drift, and model performance degradation in production. Uses PSI, KS-test, and chi-squared for statistical drift, plus evidently and nannyml for automated reports. Use when monitoring a deployed model or comparing training vs production data distributions.

Systematic evaluation of ML models, experiments, and AI system outputs. Multi-dimensional rubrics, LLM-as-judge, bias detection, and structured comparison frameworks. Use when the user asks to "evaluate model performance", "compare models", "build evaluation rubrics", "assess output quality", "detect model bias", or mentions evaluation frameworks, LLM-as-judge, model comparison, or quality assessment.

Script macOS apps via AppleScript/JXA: Mail, Calendar (EventKit attendees), Notes, Reminders, Finder (disk/file info), and Spotlight/mdfind (full-text and metadata search across the file system).

Control macOS system audio via osascript. Use when getting or setting system volume, toggling mute, listing audio input/output devices, or checking which device is the default output.

Automate Google Chrome with osascript/JXA. Use when listing/focusing/closing/navigating/reloading tabs, muting tabs, managing windows, executing JavaScript, scraping page content, extracting links, filling forms, reading localStorage/cookies/audio state, or reading Chrome history/bookmarks/downloads/extensions/profiles.

Read and write the macOS clipboard with full type support. Use when reading clipboard text, HTML, RTF, images, or file paths (Finder copy); writing rich content; detecting clipboard changes via changeCount; or intercepting content between copy and paste.

Read macOS app SQLite databases directly. Use when reading iMessage chat history, Chrome browsing history, Safari history, or any local app SQLite DB. Always copy the DB before querying — most apps lock the file while running.

Create and manage macOS LaunchAgents and LaunchDaemons. Use when setting up background automation, scheduled tasks, file watchers (WatchPaths), interval polling, or persistent daemons that survive logout or reboot.

Inspect macOS network state. Use when getting current WiFi SSID, signal strength, and channel; listing all network interfaces with IP/MAC addresses; or watching for network connect/disconnect events via Darwin notifications.

macOS notifications and system event watching. Use when sending notifications, creating actionable notifications with buttons, watching for screen lock/unlock, detecting network changes, or listening to Darwin/NSDistributedNotification system events.

Competitive intelligence for companies, products, markets, or websites. Maps sites, analyzes products and pricing, tracks news and sentiment, profiles the competitive landscape, and produces SWOT analysis. Use when analyzing competitors, evaluating markets, or auditing web presence.

Comprehensive due diligence on companies or individuals. Investigates background, product, financials, leadership, technology, risks, competitive position, and public reputation. Use when evaluating companies for investment, researching people, or making partnership decisions.

Verify claims, statements, or assertions by cross-referencing web, news, academic, and social sources. Produces a confidence-scored verdict with full evidence chain. Use when verifying claims, checking facts, or validating statements.

Comprehensive multi-source research with academic depth. Searches web, social platforms, news, academic databases, and GitHub. Discovers papers, datasets, and open source implementations. Use when conducting research, literature reviews, or investigating any topic.

Track and analyze trends, breaking news, and market shifts. Measures velocity across web, social, news, GitHub, and academic signals. Produces news digests and trend trajectory reports. Use when tracking news, identifying trends, or monitoring developments.

Stress-test any idea, proposal, or plan by finding its weakest points, probing hidden assumptions with Socratic questions, and identifying failure modes through red team analysis. Use when the user asks to stress-test, poke holes, find weaknesses, red team, challenge, or ask 'what could go wrong' or 'what am I missing'.

Build the strongest possible version of an argument, then argue the complete opposite with equal conviction. Use when the user asks to steelman something, argue both sides, make the strongest case, or understand the best version of an opposing view.

Deliver a structured verdict on any idea, argument, or proposal, identifying what survives scrutiny, what doesn't, and what needs more work. Includes the gadfly sting, the one uncomfortable truth nobody wants to say. Use when the user asks for a verdict, bottom line, final assessment, 'what survives', or 'give it to me straight'.

Full adversarial analysis of any idea, proposal, or argument. Runs steelman, crucible, and verdict in sequence for a complete challenge drawing from 10 reasoning traditions. Use when the user asks to play devil's advocate, run a full challenge, or says 'tear this apart' or 'give me the full treatment'.

Generate rich outputs from ingested sources — flashcards, quizzes, reports, slide decks, mind maps, infographics, data tables, or audio overviews. Use when the user wants to create any output from their documents.

Launch a live NotebookLM-style dashboard for managing sources and generating outputs. Use when the user wants a visual interface, dashboard, or UI for their notebook.

Ingest and vectorize PDFs or text files for RAG retrieval. Use when the user provides a PDF, document, or text file they want to analyze, study, or generate content from.

Deterministic deep research via RivalSearchMCP. 9 tools: 5-engine web search (DuckDuckGo/Bing/Yahoo/Mojeek/Wikipedia), 9-platform social search (Reddit/HN/StackOverflow/Dev.to/Medium/ProductHunt/Bluesky/Lobste.rs/Lemmy), 5-source news (Google/Bing/Guardian/GDELT/DDG), 5 academic DBs (OpenAlex/CrossRef/arXiv/PubMed/EuropePMC), GitHub search, website mapping, content extraction with OCR, and research topic synthesis. No API keys required. Use when the user needs web research, competitive analysis, content discovery, or academic paper search.

<when to delegate>

Autonomous ML experiment loop for improving NBA prediction models. Modifies autoresearch/experiment.py, runs evaluations, keeps/discards results based on NBA_CORE composite score. Use when user says "run autoresearch", "improve models", "run experiments", "optimize predictions", or "autoresearch loop".