claude-plugins

Configure the minimal single-line compact Claude Code statusline (repo, worktree, branch, model, effort, context window, context %, and session cost). Detects and resolves conflicts with an existing statusline before switching.

Show the full Claude Code session context assembled in load order. Outputs all CLAUDE.md files, auto-memory, SessionStart hook output, and skill listings in one document with source separators. Prints a summary table to stderr with scope, type, source, status, lines, tokens, and context% per source. Keywords: session context, what is loaded, show context, debug context, rules, hooks, skills.

Interactive setup wizard for git branch naming conventions. Creates or updates .claude-plugin/git-branch-naming.json with project-specific rules. Run this command to configure prefixes, ticket patterns, enforcement levels, and protected branches. Keywords: setup branch naming, configure conventions, branch config, git naming setup.

Analyze documentation health: run structural checks (broken links, orphans, duplicates, CLAUDE.md overflow, mandatory docs), then semantic analysis (README compliance, terminology, ADR completeness, content actuality). Create GitHub epic with linked issues for findings. Keywords: documentation audit, kb grooming, doc health, knowledge base.

Interactive setup wizard for fast-changing technology watchlist. Creates or updates ~/.claude/fresh-guides.json or .claude-plugin/fresh-guides.json. Run this command to configure which technologies require doc verification. Keywords: fresh-guides setup, configure watchlist, fast-changing technology, doc verification.

Display current fresh-guides watchlist configuration. Shows all watched technologies and their doc URLs. Keywords: show watchlist, fresh-guides config, current technologies.

Quick-update the fresh-guides watchlist. Usage: /fresh-guides-update add <name> <url> — add a technology, /fresh-guides-update remove <name> — remove a technology, /fresh-guides-update url <name> <url> — add URL to existing technology. Keywords: update watchlist, add technology, remove technology.

Invoked automatically when discussing a technology on the fresh-guides watchlist. Provides verification procedure: detect version, fetch official docs for that version, compare with training data, cite sources. Do NOT give version-specific advice for watched technologies without consulting this skill first. Keywords: fresh-guides verify, check docs, fast-changing technology, outdated knowledge, version check.

Generate retrospective summaries of Claude Code sessions. Use /retro session to summarize the current session (display only). Use /retro today or /retro yesterday to generate saved daily reports. Keywords: retro, retrospective, session summary, what did I do, daily report, productivity.

Interactive setup wizard for retroscope retrospective reports. Creates or updates ~/.claude/retroscope.json and .claude-plugin/retroscope.json. Run this command to configure storage directory, language, model, and options. Keywords: retroscope setup, configure retro, retroscope config, setup retrospective.

Invoked automatically before creating PlantUML diagrams to select the correct type. Covers 17 types: sequence, activity, state, class, ER, component, deployment, timing, mindmap, gantt, WBS, JSON, YAML, network, object, usecase, wireframe. Do NOT create PlantUML without consulting this guide. Keywords: diagram type, which diagram, best diagram, choose diagram, UML.

Browse and read playbook preset guidelines on demand. Invoked automatically when full reference is needed for documentation principles, GitHub workflow, git safety, debugging, or other configured guidelines. Do NOT skip this when RULES zone references it. Keywords: guidelines, documentation rules, commit checklist, ADR, PR workflow, playbook, debugging, git safety, planning, subagent verification, shell scripting, bash pitfalls, find path, strict mode.

Dump the verbatim session context from Claude's memory to a file. Writes ALL context sources exactly as they appear in the context window — no re-reading from disk, no re-executing hooks. Complementary to ctx-show (disk audit vs memory dump). Keywords: dump context, memory dump, what does Claude see, in-memory context, verbatim context.

Career direction analysis: interview + AI usage pattern mining to identify roles AI won't replace. Collects data from 8 data sources, runs adaptive interview, performs web research, and generates a structured career risk report. Keywords: ai fortune, career analysis, career risk, ai displacement, career direction, job future, ai resistant, career pivot.

Interactive setup wizard for playbook presets. Creates or updates ~/.claude/playbook.json (global) or .claude-plugin/playbook.json (project). Run this command to choose which coding guideline presets to enable. Keywords: playbook setup, configure guidelines, enable presets.

Remove the PlantUML pre-commit hook from the current project. Removes only the plantuml marker-delimited section, preserving other hook content.

Interactive setup wizard for kb-grooming plugin. Creates or updates configuration for documentation health checks: model, scope, checks, output options, and GitHub integration. Keywords: kb-grooming setup, configure doc checks, documentation audit config.

Interactive setup wizard for technology proficiency levels. Creates or updates ~/.claude/technology-explainer.json. Run this command to configure how Claude adapts explanation depth per technology. Keywords: technology explainer setup, configure proficiency, explanation depth.

Display current technology proficiency configuration. Shows all technologies grouped by level, default level, and custom sources. Keywords: show proficiency, technology levels, current config.

Quick-update a single technology's proficiency level, the default level, or add a custom source. Usage: /technology-explainer-update <tech> <level>, /technology-explainer-update default <level>, /technology-explainer-update source <tech> <url>. Keywords: update proficiency, change level, add source.

Invoked automatically when explaining expert-level technologies. Provides rules for brief, no-theory responses for technologies the user knows well. Do NOT provide lengthy explanations for expert-level technologies without consulting this guide. Keywords: expert explanation, brief answer, known technology.

Invoked automatically when explaining intermediate-level technologies. Provides rules for nuanced responses that skip basics but cover gotchas and edge cases. Do NOT explain intermediate-level technologies without consulting this guide. Keywords: intermediate explanation, nuances, gotchas, edge cases.

Invoked automatically when explaining learning-level technologies. Provides rules for detailed theory-first responses with examples and step-by-step walkthroughs. Do NOT explain learning-level technologies without consulting this guide. Keywords: learning explanation, detailed theory, tutorial, step-by-step, teach.

Invoked automatically when creating git branches to suggest proper names. Covers prefixes: feature/, bugfix/, hotfix/, release/, docs/, test/, chore/, refactor/. Do NOT create branches without following this naming guide. Keywords: branch name, create branch, naming convention, branch prefix, git branch.

Interactive setup wizard for semantic versioning enforcement. Creates or updates .claude-plugin/semver.json with version files, trigger strategy, and enforcement rules. Run this command to configure SemVer for your project. Keywords: setup semver, configure versioning, version config, semver setup, version bump config.

Invoked automatically when deciding version bumps to determine correct MAJOR/MINOR/PATCH increment. Covers SemVer 2.0.0: version format, decision tree, conflict resolution, common mistakes. Do NOT bump versions without consulting this guide for non-obvious cases. Keywords: semver, version bump, major minor patch, breaking change, version conflict, semantic versioning.

Configure the Claude Code custom statusline showing API rate limits, context window usage, git branch, and model info.

Validate that all PlantUML diagram URLs in markdown files match their source blocks. Reports missing or stale URLs and offers auto-fix.