ossian-stack

Audit, score, and improve App Store and Google Play listings — keywords, competitors, metadata, visual assets, ratings, and ASO experiments. Use for discovery, strategy, and listing audits; use asc for App Store Connect edits and gplay-cli for Play Console operations.

Settle caller usage, types, signatures, and module ownership before implementing code that crosses boundaries.

Operate App Store Connect with asc: metadata, pricing, builds, TestFlight, signing, submissions, analytics, and release automation.

Pre-commit/ship code review: Codex default; optional Claude or Pi.

Drive a GitHub or Azure DevOps PR to merge-ready, resolve feedback, or report status. Preserve the requested mode; never merge.

Accessibility engineering for UI components, modals, menus, forms, and widgets. Use for accessibility/a11y/WCAG, ARIA, focus, keyboard navigation, screen readers, alt text, touch targets, reduced motion, semantic HTML, form errors, or 'not keyboard accessible'.

Find what a change could break beyond its diff and prove the safety assumptions that matter before delivery.

Restate the last message in plain human language, with no jargon.

Manage Clerk auth, users, organizations, sessions, configuration, impersonation, webhooks, and API requests with the Clerk CLI.

Route implementation work to Codex CLI; Claude specs, reviews, verifies.

Simplify and review, then commit and push completed work directly on main or develop, or create, update, and drive a PR for other branches or when project policy requires it. Route verified merged branches and worktrees to targeted cleanup.

Turn an existing article, essay, newsletter, draft, or substantial note into native X and TikTok posts or a YouTube idea and script draft in the writer's voice. Use for editorial repurposing, not original long-form drafting or app ads.

Operate Convex deployments, functions, data, environment variables, logs, and codegen from the CLI. Not for authoring backend functions.

Write or improve marketing copy and product UI text: listings, landing pages, ads, social scripts, onboarding, and microcopy. Editorial long-form is outside this skill.

Create and prove a project-local verification skill that launches and drives the real app, CLI, or service.

Diagnose and fix bugs or performance regressions using a reproducible signal and verified root cause. Preserve diagnosis-only scope.

Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.

This skill encodes Emil Kowalski's philosophy on UI polish, component design, animation decisions, and the invisible details that make software feel great.

Create, revise, or convert HTML presentations and PPTX decks, including visual style exploration and export.

Remove one verified merged local branch and its exact worktree, or audit repository-wide Git clutter when explicitly requested. Protect long-lived branches, dirty work, and unique commits.

Manage Google Play Console builds, releases, rollouts, listings, products, pricing, testers, reviews, and app access with gplay.

Converge on a frontend look through rounds of prototypes and grilling verdicts. Use when the user wants to iterate on UI/visual taste against concrete variants.

A relentless interview to sharpen a plan or design.

A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.

Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.

Pause work at a safe boundary or resume it from a durable handoff reconciled with live repository state.

Explain code behavior, runtime flow, architecture, or ownership. Use for how/where questions; use why for historical rationale.

Automate SwiftUI iOS marketing screenshots across locales, devices, and appearances. Triggers: App Store assets, locale screenshots, widget renders, simctl, fastlane alternative.

Audit and repair a project's verification skill and feature map against source and live behavior. Explicit invocation only.

Replace metaphor and flourish with direct statement, in your own output or in text you are given. Use when writing or reviewing anything a person will read - a report, a review, an audit, a summary, a draft - and whenever the user asks for plainer, less writerly, less performed prose, or says something reads as trying too hard. Also use as a standing constraint on your own user-facing writing when a skill or workflow routes its output through it.

Work directly in Ossian's Obsidian markdown vault. Triggers: check/read my notes, capture/find/edit a note, daily note, MOC, Zettelkasten, second brain.

Design or audit mobile onboarding. Use for activation, questionnaire flows, first-session handoff, permission timing, and paywall placement.

Retrieve missing secrets and environment values securely with the 1Password op CLI. Use existing secure authentication when it already meets the task.

Audit online articles, newsletters, LinkedIn posts, and X threads for reader appeal, hooks, pacing, structure, and credibility. Compare revisions and use fresh reviewers when independent judgment matters. Generate headlines and openings when asked; research frames and select topics using reading, notes, and audience feedback. Does not draft full pieces or publish.

Apply Ossian Stack's working standards to the current request and choose the relevant workflow. Use only when the user explicitly invokes ossian-mode.

Diagnose live performance, analyze captured profiles, or improve one metric through measured experiments.

Operate the Post Queue command surfaces effectively. Use for the postq CLI, local Next.js and Convex development, production-safe reads, Convex deployment inspection, tests, lint, typecheck, builds, and scheduler verification.

Apply when you catch yourself writing the same instruction a second time, or notice a recurring correction. Encode the rule as a lint, metadata flag, runtime check, or script instead of more text.

Apply when facing a novel UI interaction or architectural decision with no precedent in the codebase. Build 2-3 competing prototypes and compare side by side before committing.

Apply when debugging. Trace each symptom to its root cause and fix it there; reproduce first, ask why until you reach it, resist nil-check guards that silence crashes.

Apply when writing stateful logic, or when code branches a lot or repeats a shape assumption across files. Encode the domain in a structure instead of scattered conditionals.

Apply after completing a task, before declaring done. Verify against the real artifact (run the feature, read the actual value, inspect the diff), not a proxy, self-report, or 'it compiles.'

Apply when integrating a new requirement into an existing design. Redesign as if the requirement had been a foundational assumption from day one, instead of bolting it on.

Apply to multi-step work (sweeps, migrations, runs of similar edits) and to how you stack commits and PRs. Break work into small units that each end in a verifiable state, check each before the next, and order delivery so the sequence proves itself to a reviewer.

Apply when sequencing an addition, refactor, or rewrite. Remove dead weight, redundant validators, and stub references first, then build on the simpler base.

Generate app privacy policies, terms, EULAs, App Store data disclosures, Privacy Nutrition Label mapping, SDK/data collection checks, region-specific legal docs.

Establish or resume one persistent project HQ that accepts the user's task queue, coordinates execution owners, and keeps durable operational state. Explicit invocation only.

Build a disposable UI or logic prototype to compare alternatives and answer a design question.

Restructure working code while preserving behavior: rename, extract, move, deduplicate, or reshape modules and APIs.

Release iOS and Expo apps to TestFlight and the App Store. Triggers: iOS release, Expo release, version bump, TestFlight, App Store review, Xcode Cloud, EAS, .ios-release.env.