Skills by drumrobot

Run a 4-stage research → plan → review → TDD-implement workflow for code changes with optional PR capture.

Manage Claude Code sessions: lookup, search, repair, migration, compression, and cleanup in one command surface.

Manage Git worktrees, ghq clones, SourceGit registration, and multi-account credentials with safety checks.

Diagnose agent mistakes, harden the prompt that caused them, and finish the original task.

Convert plans and code changes into GitHub issues and PRs, with auth handling, dependency tracking, and pre-publish scans.

Convert functionality into Claude Code automations (Agent, Skill, Rule, Hook, Command)

Manage the full lifecycle of Claude Code skills: create, lint, merge, convert, upgrade, and discover them.

Get a concrete split, squash, or message-rewrite plan for staged and unpushed Git changes.

Suggests next steps after a task finishes, with stall detection and ask gates before composing options.

Track multi-step session work as a checklist with two dispatch modes: registration or resume.

Consolidate AI PR feedback into classified findings and post an AI Review Summary.

Schema and lifecycle management for fix_plan.md and checklist.md trackers with priority triage and external state syncing.

Environment-aware browser operations. Detects wmux/cmux/Orca/tmux and routes to backend. Topics — ui-test (snapshots, click/fill, shadow DOM), credential-issue (login via backend -> wait sign-in -> issue/refresh token/secret), orca-computer (drive the user's real signed-in browser via `orca computer`, for bot-blocked sign-in). Use when: "browser", "web-browser", "ui-test", "credential-issue", "orca-computer", "orca computer", "computer use", "playwright", "chrome-devtools", "UI check", "browser test", "screen verify", "Playwright test", "shadow DOM cascade", "::part not working", "CDP trace", "issue token", "service credential", "open login screen", "PAT refresh", "scope expansion", "device-code auth", "browser device-code", "GitHub social login", "Cloudflare Turnstile", "bot-blocked login", "browser may not be secure".

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express...

AI harness for stable LLM workflows. Topics — pipeline (clarify → ground → plan → generate → verify, dispatches to code-workflow) [pipeline.md], guardrails (denylist + scope + conditional-reject, self-contained for openclaw headless) [guardrails.md], recovery (fail-analyze → adapt → fallback, self-contained) [recovery.md]. Use when enforcing stable AI agent workflows, applying guardrails to autonomous execution, or recovering from verification failures. "harness", "AI harness", "pipeline guardrails", "fail recovery", "workflow stability", "agent harness" triggers

Chocolatey operations integration — post-processing after choco upgrade, nssm service path refresh, NSSM → shawl migration (syncthing v2 etc.), resync on UniGetUI/choco metadata update failures. diagnose [diagnose.md], metadata-fix [metadata-fix.md], post-upgrade [post-upgrade.md], shawl-migration [shawl-migration.md], update-path [update-path.md]. Use when: "choco", "chocolatey", "choco upgrade", "choco outdated", "vcredist", "nssm", "nssm recovery", "nssm path", "service path refresh", "after choco upgrade", "service failure", "SERVICE_STOPPED", "shim issue", "syncthing nssm", "syncthing v2", "shawl migration", "nssm deprecation", "service-specific error", "UniGetUI", "metadata update failure", "metadata stale", ".nuspec stale", "choco metadata".

Route TODO checklists to the right storage + TaskList conversational discipline. Topics — session (/wip), file (fix_plan.md), issue (GitHub Issues), completion-report, fix-plan-sync, priority-prefix, media-separation (tracking vs recording vs knowledge). Use when: "TODO management", "checklist", "todowrite", "fix_plan cleanup", "register as issue", "completion report", "task priority", "media-separation".

Convert Markdown / HWP / DOCX / RTF to PDF, PNG, HTML, DOCX. LibreOffice is the top-priority converter (watermark-free, MPL 2.0, faithful layout, handles HWP natively). prince / md-to-pdf are internal-only fallbacks — prince stamps a non-commercial watermark, so it must not touch external deliverables. Topics — pandoc (Markdown to PDF/PNG), docx (Word + md-to-docx-to-pdf chain, delegates to the docx skill) [docx.md], legacy (HWP / DOC / RTF via LibreOffice, in this SKILL.md), marp (slides with Mermaid, checklists, overflow) [marp.md]. Rich-format analysis defaults to PDF or HTML with images/tables preserved, never a silent txt fallback. Triggers — "PDF conversion", "PNG conversion", "DOCX conversion", "Word conversion", "HWP conversion", "hwp pdf", "hwp analysis", "legacy document", "document export", "export pdf", "md to pdf", "md to docx", "docx to pdf", "marp mermaid", "official PDF", "watermark-free PDF", "LibreOffice conversion", "document analysis".

chezmoi dotfile management. apply - interactive diff review per file [apply.md], consolidate - merge duplicate templates [consolidate.md], cross-platform - Windows/macOS compatibility [cross-platform.md], doctor - check required files and copy if missing [doctor.md], mcp-sync - MCP server synchronization [mcp-sync.md]. For commit splitting → commit-splitter. Use when "chezmoi apply", "chezmoi diff", "diff review", "per-file apply", "chezmoi consolidate", "duplicate templates", "MCP sync", "chezmoi Windows", "cross platform", "utcp to mcp", "move to global mcp", "chezmoi mcp migration", "chezmoi doctor", "missing script", "SourceGit".

Claude Code & Antigravity session management. Topics — id (lookup UUID), list (enumerate), search (keyword validate), import, summarize, analyze (stats), archive (flat bak), classify, clean-profanity (sanitize JSONL), split (boundaries), compress (MCP), context (usage measurement pull/push), destroy, dual-sync (Win/WSL memory), install (hook), memory-trim (budget trim), migrate (to worktree), move (update cwd), purge (dead sessions), rename (custom title), repair (chain/tool_result/UUID), rewind (context truncate), url (web URL). Use when: "session id", "current session", "session list", "list sessions", "session search", "find session", "session classify", "session compress", "session migrate", "session move", "session repair", "session rename", "session split", "session purge", "session rewind", "session analyze", "session import", "session summarize", "session archive", "archive session", "clean profanity", "session cleanup", "memory trim", "dual sync", "measure context", "context u

Synchronization management with external tools. agents - Windows+WSL ~/.agents share + ~/.claude/plugins split layout [agents.md], chezmoi - dotfile template management [chezmoi.md], knowledge - session knowledge → Serena memory [knowledge.md], mcp - MCP server synchronization [mcp.md], syncthing - chezmoi Syncthing sync and diagnostics [syncthing.md]. Use when "knowledge sync", "chezmoi add", "dotfile management", "syncthing", "MCP server add", "MCP sync", "external sync", "(?d)", ".stignore", "ignored files", "stignore settings", "sync incomplete", "sync status", "DB reset", "stale cache", "syncthing diagnostics", "index reset", "rescan", "encryption consistency", "Failed to verify encryption", "receive-encrypted mismatch", "garbage encryptionPassword", "Antigravity Syncthing", "Gemini Syncthing re-register", "Windows WSL agents layout", "plugins split", "marketplace EACCES". For ClawHub-related tasks → delegate to /clawhub skill.

Project initialization toolkit. contributing - auto-generate CONTRIBUTING.md from project structure [contributing.md]. "init", "project init", "initialize project", "contributing guide", "CONTRIBUTING.md", "contributing generate" triggers

Claude Code plugin lifecycle management. cache - clean old cache versions (cleanup only — NOT for cache miss diagnosis) [cache.md], clustering - skill affinity scoring (coupling/dep/hook-ownership) → plugin bundle membership [clustering.md], create - plugin authoring guide (structure, plugin.json, components) [create.md], dev-reflect - reflect local dev repo into marketplace clone for pre-push testing [dev-reflect.md], hud - OMC HUD statusline (omcHud elements, omcLabel, wrapper sed fallback) [hud.md], marketplace - clone/list/update marketplace repos [marketplace.md], troubleshoot - cache miss/error, install fail, cache sync, HUD diagnostics — ALL plugin errors route here [troubleshoot.md]. "plugin", "marketplace", "plugin install", "plugin not installed", "cache miss", "cache error", "plugin error", "load error", "reload errors", "dev reflect", "plugin clustering", "bundle skills", "skill affinity", "OMC HUD", "omcHud", "omcLabel", "statusline", "[OMC#" triggers.

MCP server configuration and diagnostics. add - add server (scope selection + registration) [add.md], move - change server scope (project→local, local→user) [move.md], format - JSON format reference (differences per agent) [format.md], catalog - list of commonly used servers [catalog.md], diagnostics - connection failure diagnosis and troubleshooting [diagnostics.md]. Use when: "add MCP", "mcp-config", "MCP server config", "add context7", "MCP format", "MCP server list", "move MCP", "MCP scope", "scope change", "MCP connection failed", "MCP error", "Failed to reconnect", "code-mode not working", "MCP server issue".

Unified backlog lifecycle management and task tracking. Orchestrates session TODOs, workspace checklists (`fix_plan.md`, `checklist.md`), and issue trackers with vendor-agnostic lifecycle contracts. Topics — triage (classify incoming requests into session/file/issue backlogs), priority (GitHub-aligned P0-P3 urgency tagging and blocker triage), sync (external tracker status polling and resolution), prune (demote lower-priority backlog noise from active focus), lifecycle (authoring, state transitions, DoD criteria), comment (post follow-up notes/sub-findings to parent issues), create (issue and intake issue creation), record (session completed work and next-session candidates). Use when: "backlog", "backlog triage", "backlog sync", "backlog prune", "task lifecycle", "manage backlog", "backlog priority", "backlog cleanup", "plane backlog", "issue comment", "backlog record".

Launch and coordinate agent sessions inside the Orca IDE runtime (stablyai/orca). Topics: send - find a running Orca terminal by title/preview/worktree and deliver a prompt to it, always disambiguating with a question when more than one candidate matches and never targeting the caller's own pane [send.md]. launch - start a new agent session (claude, antigravity, openclaw, codex, and 30+ other supported agents) in a fresh or existing worktree [launch.md]. install - install the stablyai/orca skill bundle into Claude Code, OpenClaw, or Antigravity, asking whether to ghq-clone + symlink or use a plain marketplace install [install.md]. Use when: "send this to the claude session working on X", "hand this off to another agent in Orca", "spawn a new codex/antigravity worktree", "which Orca terminal is running Y", "install the orca skills into openclaw/antigravity", "Orca worktree", "Orca terminal send".

Oh My Zsh management. plugin - Add/install plugins to .zshrc plugins=() array [plugin.md], custom - Write $ZSH_CUSTOM/*.zsh based on requirements [custom.md]. Use when "omz", "oh-my-zsh", "zsh plugin", "add plugin", "custom zsh", "zshrc", ".zshrc", "zsh config", "zsh function", "zsh alias" is mentioned.

Run the self-improving loop before session end. config - enable/disable individual tasks [config.md], hook-review - review hook errors and suggest improvements [hook-review.md], rag-store - persist to RAG before session end + sync fix_plan completed items to RAG (medium matrix fallback) [rag-store.md], run - 4-step sequential execution pipeline, entry point and Steps 0-1 [run.md]; its step bodies live in run-self-improve.md (Step 2), run-knowledge-persist.md (Step 3), run-checklist-sync.md (Step 4-A/4-B) and run-wip-handoff.md (Step 4-C/Step 4.5). Mistake recording (retrospect) and FA pruning moved to the fa skill — invoke Skill("fa") / Skill("fa", "fa-prune"). Supports Ralph mode (records to improvements.md instead of AskUserQuestion). Use on "wrap up", "session cleanup", "end session", "cleanup", "record mistake", "save feedback", "improve", "retrospect", "hook error", "next action", "RAG store", "qdrant store", "fix_plan sync", "cleanup --auto", "cleanup --ralph", "auto mode", "ralp

Guide and conventions for running the ralph autonomous loop tool (frankbria/ralph-claude-code) with Claude Code. Covers wrapper setup that avoids Circuit Breaker false positives, PUBLIC repo language guards for autonomous posting, and the workspace `.ralph/` artifact layout.

GitHub repository configuration and verification. Topics — actions (GitHub Actions workflow generation + Dependabot grouping by dependency type), setup (integrations: CodeRabbit, Copilot, issue/PR templates, CODEOWNERS, branch protection), verify (repo prerequisites checklist for github-flow: remote, default branch, PR template, branch protection, GitHub Actions, CONTRIBUTING, LICENSE). Use when: "github actions", "workflow generation", "CI workflow", "dependabot", "dependabot groups", "coderabbit", "copilot setup", "GitHub Actions", "repo setup", "CODEOWNERS", "PR template", "issue template", "branch protection", "repo verify".