plugins-kit

Use when creating, listing, working, closing, archiving, or moving task folders and task_list refs. Do NOT use for native TaskCreate/TodoWrite or code review.

Use when interpreting SessionStart bootstrap messages or configuring user/project dependency manifests. Do NOT use for non-bootstrap debugging.

Use when md-audit dispatches a CLAUDE.md audit against the cohesion framework; fans multi-file runs via the Workflow tool. Do NOT use for SKILL.md.

Use when md-authoring dispatches authoring a CLAUDE.md -- a valid claude_md block (scope, insights). Do NOT use for SKILL.md (use skill-authoring).

Use when deciding where a fact should live across CLAUDE.md / SKILL.md / references (placement). Do NOT use for content shape (use md-authoring).

Use when authoring or refining an md artifact -- a SKILL.md or a CLAUDE.md -- or via /md-authoring. Do NOT use for auditing (use md-audit).

Use when md-audit dispatches a project-document audit against the cohesion framework. Do NOT use for SKILL.md, CLAUDE.md, or cross-references.

Use when auditing an md artifact -- a CLAUDE.md, a SKILL.md, a project document, or skill cross-references -- or via /md-audit. Do NOT use to author skills (use md-authoring).

Use when reading, writing, or extracting Unreal Editor data via Python. Do NOT use for blueprint edits or non-Python Editor work.

Use when the user invokes /statusline or asks to inspect, customize, or reset their Claude Code status line. Do NOT use for unrelated UI customization.

Use when md-audit dispatches a scan for broken skill cross-references; fans multi-file runs via the Workflow tool. Do NOT use for single-skill validation.

Use when authoring or refining a Claude Code skill -- types, contracts, authoring-time validation. Do NOT use to audit skills (use md-audit) or invoke them.

Use when designing a tool that produces a materialized insight from deep scans or inference over project data. Do NOT use for one-shot print-and-exit scripts.

Use when md-audit dispatches a SKILL.md audit (contract, cohesion, roster/hierarchy), fanning multi-file runs via the Workflow tool. Do NOT use for CLAUDE.md.

Use when generating or refreshing the Claude Code plugin ecosystem poster (16:9 HTML browser of marketplaces and plugins). Do NOT use for skill authoring.

Use when generating or browsing the claude-explorer view -- an HTML view of ~/.claude/ + the current project. Do NOT use for audits or skill authoring.

Use when detailed context tracking is needed (context-switching, background work). Do NOT use for standard reporting.

Use when authoring, validating, or running workflow-kit workflows. Do NOT use for one-off Workflow scripts Claude writes directly.

Use when cleaning up Unreal ObjectRedirector assets in a P4-backed UE project. Do NOT use for non-P4 projects.

Use when automating Unreal Engine -- reading/writing data, MCP editor control, or redirector cleanup. Do NOT use for C++/Blueprint work.

Use when reviewing local git changes -- before push, before opening a PR, or auditing a branch. Do NOT use for Perforce CLs or existing PRs by URL.

Use when reviewing a pending Perforce changelist, or before asking the user to submit a CL. Do NOT use for git diffs or submitted CLs.

Use when the user asks about cache hit rate, token usage, or prompt-caching stats. Do NOT use for runtime cache configuration.

Use when encoding a discovered insight into a persistent location (CLAUDE.md, skill, reference doc). Do NOT use for end-of-session doc review.

Use when ending a session to review work and recommend doc updates to CLAUDE.md, skills, and reference docs. Do NOT use for new skills or one-off notes.

Use when asked to convert an HTML file to a PDF and open it. Do NOT use for non-HTML inputs, PDF editing, or page screenshots.

Use when checking, setting, or rotating the OpenRouter API key, or diagnosing 401/402 errors. Do NOT use for general LLM/translation work.

Use when automating the Unreal Editor via the MCP server -- spawning actors, authoring graphs, driving PIE. Do NOT use for on-disk asset inspection.