Studio-Moser
- 25 skills
- 0 followers
- 5 hours ago last updated
- ▌ Setup 2 · studio-moserUse when Harness has been installed but this developer still needs a personal agents-repo relationship, portable links, runtime capability discovery, a model rubric, or optional Shelby enrichment configured.
- ▌ Setup 3 · studio-moserUse when Product Pulse needs onboarding or reconfiguration for a project's research context, sources, operational config, and report directories.
- ▌ Setup · studio-moserUse when PM has not yet been configured for a workspace, or when the user explicitly requests reconfiguration of its issue-tracker backend.
- ▌ Ingest · studio-moserUse when new Product Pulse daily, weekly, or deep-dive reports need importing into the configured issue tracker. Do not use to triage existing candidates or reconcile completed work.
- ▌ Triage · studio-moserUse when `status/needs-triage` items require keep, reject, or deduplication decisions or readiness preparation before execution. Do not use for raw-report ingestion, ready-item implementation, or tracker reconciliation.
- ▌ Dev Task · studio-moserUse when implementing one named feature, known bug fix, or focused code change interactively in the current repo, especially when a teammate needs guided approval gates. Do not use for a sprint or backlog batch, open-ended design, standalone review, or diagnosis before a cause is known.
- ▌ Sync · studio-moserMake this machine match your personal agent repo — the private repo holding your skills, global CLAUDE.md, and shared Claude Code settings. On first run it clones an existing private repo or safely creates one from loose configuration; after that it ingests the current remote before reconciliation, then commits and pushes this machine's changes. Re-links anything that drifted back into ~/.claude, lints for paths that would be wrong on another machine, and optionally triggers a pull on your other machines. Trigger: "sync my config", "sync my machines", "update my skills from my repo", "is this machine up to date", or /harness:sync. Do NOT use as the first-time user-facing setup workflow (that's /harness:setup), for creating the model rubric (that's /harness:model-rubric), or for anything in a project repo — sync only touches this developer's user-global agent config. Harness setup invokes this skill internally for its existing repository, link, and portability mechanics.
- ▌ Reconcile · studio-moserUse when the issue tracker may be stale after merges or sprints, or needs a periodic reality check against git history and current project state. Do not use for deciding new items or implementing ready work.
- ▌ Sprint Dev · studio-moserUse when the user wants to build one or more already-ready `owner/ai` backlog items from a configured tracker. Do not use for one named change, an untriaged item, or tracker reconciliation.
- ▌ Review · studio-moserUse when delegating or performing a review of a frozen, pinned target — a commit SHA, tag, or immutable plan/patch snapshot — through Harness routing. Covers: dispatching the configured review route; running or requesting an independent, fresh-context, cross-provider, or adversarial review (which needs explicit cost approval first); reviewing work authored by another model or worker so the reviewer stays independent of the author; and natively re-verifying, reproducing, confirming, or withdrawing findings that an external or sub-agent reviewer claimed against that fixed target. Also use to return a typed blocked or accepted HarnessResult for such a request. Do not use for ordinary review of a moving working tree or uncommitted diff, replying to or fixing GitHub PR comments, visual or browser-based UI review, or evaluating prompts, rubrics, vendors, or documents.
- ▌ House Rules · studio-moserUse when a code change needs Studio Moser conventions for change classes, branches, file naming, commits, pull requests, testing, or pre-commit security checks.
- ▌ Execute · studio-moserUse when a workflow has a bounded Harness implementation request that should run through a resolved semantic route, whether the available executor is the native agent runtime or an internal cross-provider adapter.
- ▌ Generate · studio-moser bundleGenerate images, video, music, or speech through the Kie.ai MCP with a budget guard, local archiving, and a prompt log. Prices the batch and stops for confirmation before spending, downloads every result to a dated folder, and records the prompt that produced each file. Use for ad creative, brand imagery, illustrations, b-roll, image-to-video, upscales, or any batch of generated media. Invoke with /generate:generate.
- ▌ Audit · studio-moserAudit existing screenshot captures for quality issues. Checks every PNG file in a directory for wrong-company content, blank sections, cookie banners, error pages, and other problems. Reports what needs recapturing. Use when screenshots look wrong or after a batch capture to verify quality. Invoke with /site-capture:audit.
- ▌ Computer Use · studio-moserUse when a Harness request needs proof from driving a machine, not reading code: launch or attach to a live app, page, browser, simulator, or device; exercise the flow; and capture fresh visual or runtime evidence you inspect. Trigger for opening, clicking through, verifying on screen/device, or proving behavior from a worktree under stated path, permission, or artifact constraints. Also use when Accessibility, Screen Recording, or simulator capability may be missing, so the result must be blocked rather than replaced with code inspection. Skip reading existing screenshots, one-off screenshot capture without workflow verification, writing tests without running a browser, explaining OS settings, or code-only verification.
- ▌ Model Rubric · studio-moser bundleCreate or refresh this developer's user-global model-routing rubric — the file that decides which model does which work (cheap models for bulk/mechanical work, the strongest for ambiguous or taste-sensitive work). Lives at ${XDG_CONFIG_HOME:-$HOME/.config}/studio-moser/model-rubric.yml, one per developer; on machines with an agents repo the folder is a symlink into it, so the rubric syncs across machines. Trigger: "set up my model rubric", "refresh my rubric", "which model should agents use", "my rubric is stale", or /harness:model-rubric. Do NOT use to route a specific task right now (just read the rubric), or to configure a project's issue tracker (that's /pm:setup). Harness setup invokes this skill internally after discovering the current machine's capabilities.
- ▌ Capture · studio-moserCapture full-page screenshots of one or more websites. Handles scroll-triggered animations, lazy-loaded content, cookie banners, and bot-protected sites. Uses a tiered approach: Microlink API first, Chrome browser fallback for protected sites. Use when you need design references, competitive analysis screenshots, or visual audits. Invoke with /site-capture:capture.
- ▌ Transcribe · studio-moser bundleFetch a spoken-word transcript from a video URL. Supports YouTube, YouTube Shorts, Instagram posts/Reels, TikTok, and Threads. Use whenever you need the words said in a video — for research, summarization, or analysis. Invoke with /transcribe:transcribe <url> or from Bash as `transcribe <url>`.
- ▌ Feature Walkthrough · studio-moserUse when a developer asks to see, demonstrate, record, or visually review a web feature through an existing browser-testing workflow.
- ▌ Deep Dive · studio-moser bundleUse when the user explicitly asks to research, analyze, or compare external videos, articles, repositories, documentation, or code against the current project.
- ▌ Mac App Design · studio-moser bundleUse when designing, building, or reviewing the interface of a native-feeling macOS desktop app implemented with Tauri, Svelte, HTML/CSS, Rust, or a WebView; especially when it resembles a website, dashboard, enlarged mobile app, or imitation Mac window. Not for Windows, Linux, iOS, Android, or SwiftUI-first UI.
- ▌ Daily Research · studio-moser bundleUse when configured research domains need a daily source scan, strategic filtering, and a dated report for PM ingestion and publication.
- ▌ Designing In Figma · studio-moser bundleUse when authoring, generating, or building a design, screen, UI, mockup, or component DIRECTLY INTO Figma via the Figma MCP (the use_figma / write-to-canvas path) — i.e. code-to-design, not design-to-code. Triggers: "design this in Figma", "build the screen in Figma", "make a mockup in Figma", "create a Figma component", "generate a UI into Figma". Also use when moving a design system INTO Figma from any source — a Claude Design project, a Storybook, CSS custom properties, a component library, design tokens. Use especially when Figma output looks more generic than what Claude produces in HTML/CSS — this skill closes that gap. Do NOT use for pulling existing Figma designs into code (that is design-to-code).
- ▌ Weekly Strategist · studio-moser bundleUse when the last 7 daily reports and current market evidence need a weekly strategy brief, exactly three priorities, and recommendations for PM ingestion.
- ▌ Design Critic Loop · studio-moser bundleUse to raise a design to a high bar — a website, screen, UI, presentation, graphic, mockup, HTML/CSS or SwiftUI/AppKit output — via the Gauntlet Loop: independent fresh-context critics that judge against an explicit "done" bar and iterate until they pass, instead of the generator grading its own homework. Triggers: "critique this design", "run the gauntlet", "design loop", "polish this", "make it better", "raise the bar", "is this good enough". Also invoke it from another design skill after building, to verify the output against that skill's checklist. If "done" is unclear, it asks the user for goals and reference examples first. Reserve for high-value / reusable artifacts — the loop is token-expensive. NOT for routine one-off tweaks.