porkytheblack

Expert guide for building AI-powered applications with the Glove framework. Use when working with glove-core, glove-react, glove-next, tools, display stack, model adapters, stores, or any Glove example project.

Expert guide for integrating Oasis update server with Tauri apps for auto-updates, crash reporting, and feedback collection.

Expert guide for integrating Oasis into Tauri applications. Use when working with oasis-sdk, auto-updates, crash reporting, user feedback, release workflows, or any Oasis-powered Tauri project.

Expert guide for setting up Tauri deployment pipelines with GitHub Actions, code signing, and Oasis update server integration.

Expert guide for bumping versions, creating git tags, and managing releases for Tauri and Node.js projects.

Guide for downstream coding agents driving Glorp — especially Glorp Garage (the remote multi-session control plane) and its sandboxes (Garage-provisioned workspaces). Use when programmatically creating sessions, running coding agents over the REST/WebSocket API, the @porkytheblack/glorp-client SDK, or the @porkytheblack/glorp-mcp MCP server, provisioning tenant namespaces, sandboxes from templates, exchanging files, or wiring Glorp into CI/orchestration. Also use for the higher-level Task API — the black-box surface where you submit a typed job (make a video, build a deck, review a PR, fix a bug) and poll one object for the result, without managing sessions, templates, or the event stream.

Expert guide for building AI-powered applications with the Glove framework. Use when working with glove-core, glove-react, glove-next, tools, display stack, model adapters, stores, any Glove example project, or deploying agents as sandboxed runtime services with Glovebox (glovebox / glovebox-kit / glovebox-client).

Enforce NASA-style "Power of Ten" software engineering discipline on any code being written or modified, with a hard guideline that no source file exceeds 200 lines so a reviewer can read it top-to-bottom in one editor screen without jumping around. Use this skill whenever writing new source files, refactoring existing ones, reviewing code, splitting modules, or setting up a new project — even if the user does not explicitly say "under 200" or "NASA". If the user mentions file length, file size, "this file is too long", complexity, readability, code-review friendliness, or asks to set up linters/formatters/CI, apply this skill. Also apply proactively whenever any source file being edited is approaching or exceeds 200 lines.

Use this skill when building with the Station background job framework. This includes creating signals (background jobs), defining broadcasts (DAG workflows), authoring runtime-editable dynamic broadcasts, scheduling signals/broadcasts at runtime, writing expressions for `input` mappings and `when` guards, configuring adapters (SQLite, PostgreSQL, MySQL, Redis), customizing API key storage, setting up runners, writing subscribers, and configuring the Station dashboard. Station is a TypeScript-first framework for type-safe background jobs with Zod validation.

Expert guide for building AI-powered applications with the Glove framework. Use when working with glove-core, glove-react, glove-next, tools, display stack, model adapters, stores, any Glove example project, or deploying agents as sandboxed runtime services with Glovebox (glovebox / glovebox-kit / glovebox-client).

Use when integrating the npm packages `mailbox-station` (provider-agnostic mailbox-watch core) or `gmail-station` (Gmail provider). Triggers for: wiring up `createStation`, writing or porting a `StoreAdapter`, building a `MessageHandlerFn`, configuring the Gmail provider (OAuth, Pub/Sub, watch lifecycle), tuning the trigger worker (concurrency, backoff, dead-letter), interpreting tagged errors, scheduling watch renewal, or any question of the form "how do I do X with mail-station / gmail-station". Do NOT trigger for: generic email parsing libraries, raw Gmail API usage outside this stack, IMAP, or sending mail.