dotfiles

Write messages, comments, PR reviews, Slack posts, and other prose in Sean's voice. Use when drafting anything that will be sent under Sean's name — code review comments, postmortem feedback, Slack messages, GitHub comments, coaching notes, design feedback, strategic takes. Triggers: 'write like me', 'in my voice', 'draft a comment for me', 'sound like me', 'sharpen this', 'this still sounds like AI'.

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/PRD asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".

Implement a piece of work based on a spec or set of tickets.

Plan a huge chunk of work — more than one agent session can hold — as a shared map of investigation tickets, and resolve them one at a time until the way to the destination is clear.

Move todos through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs. Maps to pi's built-in todo tool; see ../todo-tracker.md for the full tracker → tool mapping.

Turn the current conversation into a spec and publish it as a todo — no interview, just synthesis of what you've already discussed.

Break a plan, spec, or the current conversation into a set of tracer-bullet todos, each declaring its blocking edges — create-then-wire so every blocker id is real.

Run an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for a "deep code review", "strict maintainability review", or "review this branch harshly".

Guide to Jujutsu (jj) version control system. Use when working with commits, branches, pull requests, rebasing, or pushing, or when the user mentions jj, git, or version control.

Relentless codebase analysis that classifies every recurring pattern into three buckets — positive (reinforce), negative (fix), inbetween (decide) — and produces a portable artifact with migration paths. Use when asked to run 'jungle-book', 'analyze the codebase', 'assess codebase health', or 'find friction' — any honest, opinionated evaluation of codebase patterns and conventions. Works on whole codebases by default; accepts an optional scope.

Discover project conventions and surface security concerns. Use when starting work in a new or unfamiliar project, when asked to "learn the codebase" or "what are the conventions", or to check for "anything shady in this codebase".

Captures and retrieves project knowledge — decisions made, patterns learned, gotchas discovered, and architectural context in a .notes/ directory. Use when recording a decision, capturing a learning, documenting a pattern, noting a gotcha, or retrieving past project context. Triggers: 'record decision', 'capture learning', 'document pattern', 'add to gotchas', 'what did we decide about', 'project notes'.

Obsidian CLI patterns for vault operations. Use when creating notes, setting properties, searching, querying Bases, managing links, or running eval commands against the vault. Triggers: 'obsidian cli', 'vault command', 'run obsidian command'.

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

Verify implementation against declared integration contracts. Use when checking that code matches API schemas, DB schemas, event formats, or other contract surfaces. Triggers: 'verify integration', 'check contracts', 'does this match the API'.

Add an MCP server to pi. Use when asked to "add", "configure", or "register" an MCP server. Handles both global and project-local configurations.

Change-mode review procedure for the adversarial-reviewer agent. Fires when the target is a code diff, GitHub PR URL, or working-tree diff. Invoked by the agent's turn-1 dispatch.

Research-mode review procedure for adversarial-reviewer. Fires when the target is a `.md`/`.txt` file, URL, or raw claim text. Invoked by the agent's turn-1 dispatch.

Manage terminal sessions via cmux — spawn workspaces for dev servers, test runners, and background tasks. Read output, send commands, and orchestrate multi-terminal workflows.

Efficiently read and analyze pi agent session JSONL files. Use when asked to "read a session", "analyze a session", "what happened in this session", or given a .jsonl session file path.

Grilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (CONTEXT.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions, or during spec/planning when domain clarity matters.

Guide to Jujutsu (jj) version control system. Use when working with commits, branches, pull requests, PRs, version control, rebasing, pushing, or when the user mentions jj, git, or version control operations.

Read this skill before making commits

Verify that cited sources actually say what the document claims. Use when asked to "shepardize", "verify citations", "check sources", "citation integrity check", "do the sources support this", "check source fidelity", or before delivering research output. Reusable by adversarial-reviewer, researcher, and planner agents. Called as subroutine S6 by adversarial-review-research.

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Create and maintain AGENTS.md files that give AI coding agents the non-discoverable context they need to work effectively in a codebase. Use when initializing a new project for AI agents, creating AGENTS.md files, auditing existing AGENTS.md files, or when the user runs /init. Triggers: 'create agents.md', 'init agents', 'set up for AI agents', 'what should go in agents.md', 'audit agents.md', '/init'.

Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.

Iterate on a PR until CI passes. Use when you need to fix CI failures, address review feedback, or continuously push fixes until all checks are green. Automates the feedback-fix-push-wait cycle.

Capture a web article from a URL into an Obsidian vault as an EMS Article entity note: fetch page, extract main body, create note via Obsidian CLI with article properties (`is`, `url`, `title`, `author`, `publisher`, `published`, `date`). Use when user says 'save this article', provides a URL to download, or wants the body saved as Markdown.

Obsidian vault conventions for this repo (EMS/Dataview/Templater). Use when creating or editing notes or when referencing vault files in responses: always format vault file references as Obsidian wikilinks (not inline code file paths), follow frontmatter patterns, and store/consult the canonical conventions note in the vault.

Research specialist for external documentation, library APIs, GitHub repos, and technical references. Use when you need to understand how a library works, find examples of a pattern, check compatibility, look up API docs, or gather information before making a decision. Triggers: 'look up', 'how does X library work', 'find examples of', 'check docs for', 'what does the API for X look like', 'research', 'is there a library for'.

Build native macOS desktop apps with HTML/CSS/JS UIs rendered in WKWebView. Use when building or modifying apps in a ~/.valet/apps/<app>/ directory, or when the user mentions App Valet, valet new, valet build, etc. Triggers: 'valet', 'app valet', 'desktop app', 'macOS app', 'WKWebView app', 'native app from HTML'.

Captures and retrieves project knowledge — decisions made, patterns learned, gotchas discovered, and architectural context in a .notes/ directory. Use when recording a decision, capturing a learning, documenting a pattern, noting a gotcha, or retrieving past project context. Triggers: 'record decision', 'capture learning', 'what did we decide about', 'project notes', 'add to gotchas', 'document pattern', 'note this', 'remember that', 'log this decision'.

Capture a web article from a URL into an Obsidian vault as an EMS Article entity note: fetch page, extract main body, create note via Obsidian CLI with article properties (`is`, `url`, `title`, `author`, `publisher`, `published`, `date`). Use when user says 'save this article', provides a URL to download, or wants the body saved as Markdown.

Obsidian CLI patterns for vault operations. Use when creating notes, setting properties, searching, querying Bases, managing links, or running eval commands against the vault. Triggers: 'obsidian cli', 'vault command', 'run obsidian command', 'create note', 'set property', 'search vault', 'query base'.