PAS

Use when creating a local agent for a pas-clief workspace. Walks the user through naming the agent, defining its role, listing its skills, and writing CLAUDE.md and CONTEXT.md from templates. The agent lands in `library/agents/<name>/`.

Use when drafting and publishing a newsletter issue. Walks research → draft → publish phases, using the writer agent. Example library process for pas-clief — copy or extend it for your own use.

Use when applying a workspace's drafting conventions to written output — voice, format, sentence rhythm, the things that make output feel like it belongs to this project. Reads the workspace's style notes if any and adjusts the draft accordingly.

Use when line-editing a draft for clarity and economy. Tightens sentences, removes filler, rewrites awkward constructions, ensures every sentence earns its keep. Apply this skill on the final pass before output.

Use when an agent needs to gather sources or background information from the web for a topic. Produces a concise sources file with URLs, summaries, and the most relevant excerpts.

Use when checking that a pas-clief workspace is current with the latest plugin conventions, or when migrating from an older version. Walks a checklist of structural and template changes and offers to apply each.

Use when rendering a process flow as a markdown diagram — phases, agents, sub-processes, and the data passed between them. Reads the process recipe and outputs a Mermaid-compatible flowchart.

Use when creating a local process for a pas-clief workspace — a recipe that composes agents, skills, and phases into a slash-invokable workflow. The process lands in `.claude/skills/<name>/` so it becomes invocable as `/<name>`.

Use when creating a local skill for a pas-clief workspace — a callable capability that agents and processes invoke by name. Walks the user through naming, when-to-use, how-to-use, inputs, and outputs. The skill lands in `library/skills/<name>/`.

Use when scaffolding a new pas-clief workspace. Brainstorms with the user about what they're building (or what they want to figure out), then writes a tailored `CLAUDE.md` and minimum-viable structure. The deliverable is shared understanding — by the end of the conversation, the LLM understands what the user wants and the workspace reflects it.

Use when creating, managing, or running processes, agents, and skills using the script-free folder-as-workspace method. Single entry point for pas-clief — routes to the right shortcut based on user intent.

Decision guide for selecting the right orchestration pattern for a process

Evolve the PAS plugin — choose between full multi-agent cycles, quick superpowers-driven cycles, or resume a previous cycle.

Use when reviewing and applying accumulated feedback from backlogs across PAS artifacts. Invoked by the PAS router when user mentions feedback, upgrade, or improvement.

Use when creating or editing an agent within a PAS process. Usually invoked by creating-processes, not directly by users.

Use when creating or editing Claude Code hooks for a PAS process, plugin, or component. Invoked by creating-processes or standalone for adding hooks to existing artifacts.

Use when creating a new PAS process from a user's goal description. Invoked by the PAS router, not directly by users.

Use when creating or editing a composable skill within a PAS agent or process. Usually invoked by creating-agents, not directly by users.

Scan a PAS project and fix any gaps between its current state and what the installed plugin version expects.

Use when engaging with contributors on GitHub issues. Guides tone, comment structure, and when to ask for clarification vs acknowledge vs resolve.

Use when triaging open GitHub issues for the PAS framework. Reads issues, classifies them, assesses priority, and produces a triage report for Discovery.

Use in the Release phase to commit, branch, PR, and sync. Only plugin files go to main — everything else stays on dev.

Use when evaluating PAS from a new user's perspective. Assesses onboarding path, naming clarity, documentation gaps, and overall developer experience.

Use when researching Claude Code's current capabilities, competing frameworks, and ecosystem trends. Produces an opportunity report for Discovery.

Use when processing accumulated feedback signals across PAS artifacts. Scans backlog directories, classifies signals, identifies patterns, and produces a prioritized report.

Use when evaluating PAS's current state at the start of a development cycle. Audits artifacts against conventions, identifies architectural gaps, and compares capabilities to goals.

Use when producing a scoped implementation plan from approved priorities. Creates file-level specificity with parallelization analysis.

Use when validating implemented changes against the plan, PAS conventions, cross-artifact consistency, and regressions.

Scaffold a new user-controlled marketplace repository for PAS plugins. Invoked by /pas when cwd is not inside a marketplace and the user chooses to bootstrap one here.

Use when creating, managing, or improving processes, agents, and skills. The single entry point for the PAS framework.

Use when classifying user messages at process gates. Carried by orchestrator agents to determine how to handle user responses.

Use at agent shutdown to write structured improvement signals. Carried by all agents when feedback is enabled. Zero cost during productive work.

Use when classifying user messages at process gates. Carried by orchestrator agents to determine how to handle user responses.

Decision guide for selecting the right orchestration pattern for a process

Use at agent shutdown to write structured improvement signals. Carried by all agents when feedback is enabled. Zero cost during productive work.

Use when visualizing a process structure as HTML. Runs generate-overview.sh to produce a self-contained overview page.

Use when visualizing a process structure as HTML. Runs generate-overview.sh to produce a self-contained overview page.