Orkas-AI

Design review layer for OrkasVideoStudio COMPOSE previews and drafts. Review every immutable snapshot frame before showing a visual preview; when preview is skipped, use the draft as the fallback evidence. Return one complete, actionable repair set without…

Reference-media and design-system input layer for OrkasVideoStudio. Treat reference images, videos, screenshots, brand guides, and design notes uniformly, compiling reproduce/edit/guide intent into executable spatial and temporal constraints.

Aesthetic direction for OrkasVideoStudio HTML and motion-graphics compositions. Use before stage-compose writes composition-manifest.json art_direction and index.html to choose a subject-specific visual point of view, type, palette, layout signature,…

Canonical VideoStudio review authorization and state-transition policy. Use after any Gate B/C/Preview/D decision, post-gate revision, resumed approval, or exhausted visual-QA result across COMPOSE/AUTO/GENERATE/EDIT; maps explicit user authority and durable…

Deterministically assemble an approved cross-modal EDL (plan.json) into a finished video — produce each segment (edit/compose/generate/provided, delegated to its line), then assemble in ffmpeg tiers: concat primaries → overlay composed layers → mix narration…

Authoring knowledge for Orkas/OVS HTML video compositions -- write an index.html, drive animation from a paused timeline, declare canvas + duration, then run the VideoStudio draft gate to render an mp4. Trigger for explainer / animation / motion-graphics /…

The decision layer for real footage — understand → select → produce an EVIDENCE-bearing rough cut. Trigger when the EDIT task is "find / select / reduce / clean" (remove dead air, drop fillers, pick highlights, cut 1 hour to 3 minutes), NOT executing a known…

Intelligent editing of real user-supplied footage—understand it with transcript/OCR/scene/silence/quality/vision evidence, then choose deterministic timeline operations or a constrained semantic AI edit. Trigger for repurpose, montage, cleanup, localization,…

Capture, organize, search, and synthesize knowledge inside a Notion workspace. Use this skill whenever the user wants to save content to Notion, update a Notion knowledge base, or research a topic from internal Notion pages.

Work with Obsidian vaults and Web Clipper templates. Use this skill whenever the user asks to search, create, edit, move, rename, or delete Obsidian notes, maintain wikilinks, inspect vault paths, or generate an Obsidian Web Clipper JSON template.

Fetch public social posts and analyze social/campaign metrics. Use this skill whenever the user asks to collect public social discussion, analyze social sentiment inputs, calculate engagement/CTR/ROI, compare platforms/content, or turn social samples into…

Plan e-commerce merchandise creatives: visual copy, product image prompts, product video prompts, ad scripts, storyboards, CTAs, compliance gates, and creative tests.

Write and optimize e-commerce merchandise listing pages, titles, bullets, descriptions, SEO keywords, FAQs, Amazon A+ content, and competitor page diagnostics.

Research e-commerce merchandise opportunities, category viability, price bands, competitors, profit assumptions, sourcing risks, and evidence gaps across Amazon and Chinese commerce platforms.

Analyze e-commerce product reviews, ratings, and competitor feedback with privacy handling, VOC taxonomy, root-cause classification, evidence snippets, and prioritized improvement actions.

Use for product requirement analysis before PRD writing: validate product or startup ideas, synthesize user research, compare competitors, score opportunity strength, surface hidden assumptions, and decide the next validation step. Use when the user asks…

Use for repository-aware product engineering from an accepted requirement to a focused implementation and evidence-backed handoff. Covers repository intake, architecture decisions and spikes, implementation planning, feature work, bug repair, refactoring,…

Use GitHub CLI for basic GitHub operations: inspect issues, pull requests, CI runs, releases, and GitHub API data. Use this whenever the user asks to list, view, create, update, or check GitHub repository data.

Use to turn a PRD section, user story, feature slice, or behavior description into product test scenarios and pass/fail conditions. Use when the user asks for product testing, acceptance criteria, Given/When/Then scenarios, QA-ready scenarios, pass/fail…

Plan, research, draft, revise, humanize, adapt, and audit editorial content with a writing brief, source and claim ledgers, channel-aware structure, voice preservation, citation checks, and publication-readiness gates. Use for social posts, articles, blogs,…

Use for SwiftUI development work: review and refactor SwiftUI view structure, Observation usage, MV patterns, dependency boundaries, and ViewModel choices; audit SwiftUI rendering, state updates, scrolling, layout, and Instruments evidence; or use…

Use to design and implement product UI from PRD, product goals, user flows, or existing app context. Use when the user asks to build or improve a UI, frontend screen, landing page, dashboard, component set, design-system-based interface, Figma-style brief…

Author or edit a custom agent via the `<agent>` container; For: 'crystallize this conversation into an agent', 'make me an agent that does X', 'change X's workflow to ...', 'tighten this agent description'; Triggers: create agent, make an agent, crystallize,…

Author or edit a user-explicit custom skill via `<skill>` containers, metadata tags, and `<<<skill-file>>>` blocks; For: 'make a skill that does X', 'tweak the X skill', 'import this existing SKILL.md / skill package as a custom skill'; Triggers: create…