Development

Build a complete vertical slice in Protean - the full command flow from command definition through handler, aggregate mutation, event raising, and event handler reaction. This is the most common workflow for adding a new feature. Use when the user asks to "add a use case", "add a feature", "create a command flow", "implement a vertical slice", "add a new action", "build an end-to-end flow", "add command processing", or when they describe actions like "users should be able to place orders", "a...

Build a complete external integration flow in Protean using subscribers. Subscribers consume messages from external systems via broker streams and translate them into domain actions using the anti-corruption layer (ACL) pattern. Use when the user asks to "add an external integration", "consume external events", "handle webhook events", "integrate with an external system", "add a subscriber flow", "build an anti-corruption layer", "translate external messages", "process broker messages", or wh...

Build a complete read model in Protean - a projection populated by a projector that reacts to domain events. Covers single-aggregate projections (flattened view of one aggregate), cross-aggregate projections (combining data from multiple aggregates), and choosing between database-backed and cache-backed storage. Use when the user asks to "add a read model", "create a projection with projector", "build a query view", "add a CQRS read side", "create a denormalized view", "build a reporting view...

Add a complete event flow to an existing aggregate - creates a domain event, adds or updates an aggregate method that raises the event via self.raise_(), and wires an event handler that processes the event and performs specified side effects (such as syncing state across aggregates, sending notifications, or triggering downstream processes). This is the primary workflow for adding reactive, event-driven behavior to a Protean domain. Use when the user wants to "add an event", "add a domain eve...

Build a domain service flow in Protean for business logic that spans two or more aggregates. Domain services are stateless, encapsulate cross-aggregate business rules, and run pre/post invariants for validation. They are used when business logic doesn't naturally belong to a single aggregate. Wire them into command handlers or application services for persistence. Use when the user asks to "add a domain service", "implement cross-aggregate logic", "coordinate multiple aggregates", "add a busi...

Add a complete command flow to an existing aggregate - creates a command class, a FastAPI API endpoint, and a command handler with repository logic. This is the primary workflow for adding write operations to a Protean domain. Use when the user wants to "add a command", "add an action", "add an operation", "expose a write endpoint", "create a command flow", "wire up a POST endpoint", "add a new operation to an aggregate", or describes an action like "I want users to be able to [verb] a [thing...

Review a coop pull request or local diff with independent, self-validated correctness, design, convention, security, API, test, documentation, and comment lenses. Use for PR review, /review follow-up, or when asked to inspect a branch without modifying it.

Run cargo-mutants for changed coop logic and keep .cargo/mutants.toml synchronized. Use when logic-dense modules change, before refactors, or when asked to verify mutation coverage.

Run and interpret coop's VM integration suite locally on Lima or remotely on Firecracker. Use when asked for integration testing or when guest-visible/lifecycle work needs its pre-merge gate.

Run the final scope-controlled review before committing, pushing, or opening a PR. Use for closeout review, readiness checks, or self-review of a completed branch.

Shepherd the current user's open PR through base updates, CI failures, and review feedback without rewriting history or merging. Use when asked to babysit, monitor, or fix an existing PR.

Build a throwaway HTML or in-app prototype that answers one design question. Use when the user wants to try a state model or UI idea before committing, or says "mock it up", "prototype", "what should this look like".

Use for conf/system-files.toml, tracked fragments under system/, or the task and resource that merge administrator-owned files below /etc. Not for user symlinks, systemd unit enablement, or Windows registry values.

Use for the Quickshell QML/JavaScript UI and Python helper under symlinks/config/quickshell/: bars, popups, workspace layout, shared state, and network actions. Not for Rust CLI output, shell bootstrap, or unrelated application configuration.

Set up, run and respect a machine-checked definition of done. Use when the user mentions a definition of done, finish-line or commit/push gates, "the agent said done but it wasn't", wants to audit what a repo enforces, wire a PreToolUse or pre-commit or CI gate, verify a patch before it touches the working tree, collect evidence files, or fix drift between CLAUDE.md, AGENTS.md, Cursor and Copilot instruction files. Also use before reporting work finished in a repo that has a .skillgate/done.y...

Work labelled Jira tickets unattended with idle Claude Code subscription quota — set up, preview, schedule, and report on a `burn process jira` run. Use when the user says "work my Jira tickets while I'm out", "auto pick labelled tickets", "burn my backlog", "run this over the weekend / overnight / while I'm OOO", or asks what happened during an unattended burn run.

Cut, normalize, caption, and inspect short-form video and audio with ffmpeg, or analyze a reel from its actual media. Use when a clip, recording, or reel URL is supplied and the deliverable is an edit or an analysis of the footage. Writing captions or posts needs no skill.

Build or repair a Codex pet sprite atlas with the deterministic assembly and QA pipeline. Use only for Codex pet assets; general image or media work needs no skill.

Diagnose and fix code-graph index issues. Use when: search returns unexpected/empty results, or after major codebase restructuring. Covers the health check and both rebuild paths from the CLI.

Understand code structure efficiently using the AST index. Use BEFORE reading files one by one — when starting work in unfamiliar code, exploring a module before changes, or finding the right file to edit. One overview call replaces 5+ Read calls and saves significant context.

This skill should be used when the user asks to "commit changes", "create a pull request", "rebase safely", "manage branches", "fix merge conflicts", "undo a commit", "comment on a PR", "create a release", or performs any git/GitHub operations. Provides safety patterns, atomic commit formatting, PR descriptions, sensitive file detection, and GitHub API usage.

Add PostHog MCP analytics to a TypeScript, JavaScript, or Python MCP server. Instruments the server so every tool call, agent intent, and failure is captured as a $mcp_* event — the @posthog/mcp Node SDK for TS/JS, or posthog.mcp (shipped inside the posthog package) for Python. Detects the server style (official MCP SDK Server/McpServer/FastMCP, mcp-handler, @rekog/mcp-nest, jlowin fastmcp, or a custom HTTP/edge dispatcher) and wires in the matching instrumentation, credentials, and graceful ...

Use this skill when the user wants to build tool/scripts or achieve a task where using data from the Hugging Face API would help. This is especially useful when chaining or combining API calls or the task will be repeated/automated. This Skill creates a reusable script to fetch, enrich or process data.

This skill should be used when users want to run any workload on Hugging Face Jobs infrastructure. Covers UV scripts, Docker-based jobs, hardware selection, cost estimation, authentication with tokens, secrets management, timeout configuration, and result persistence. Designed for general-purpose compute workloads including data processing, inference, experiments, batch jobs, and any Python-based tasks. Should be invoked for tasks involving cloud compute, GPU workloads, or when users mention ...