Rules for isolated handler subagents spawned by clawdbot's pr-manager to own a single PR event end-to-end (review_comments or ci_failed). Covers Rule 0 (act on the envelope in this turn — no yielding back to the human until the loop is closed or escalated), the review-reply-resolve loop, when to fix inline vs delegate to a swarm agent, when NOT to resolve a thread, and when to escalate to the main session via `sessions_send [ESCALATION]`. Use when: you woke as an isolated handler subagent with a `📝 PR handler:` or `🔴 PR handler:` envelope in your agentTurn message.
Use when creating complete Claude Code plugin packages with manifests, agents, skills, hooks, and marketplaces. Use when user says "create plugin", "new plugin", "scaffold plugin", "build marketplace", "package skills".
Autonomously create, test, and validate a data connector for any web platform — end to end. Use when asked to "auto-create a connector", "automatically build a connector", or when a connector needs to be created from scratch and tested without manual guidance.
- 📄 examples.md
- 📄 prompt.md
- 📄 SKILL.md
Defines product requirements using the 5W2H framework with psychology-enhanced audience analysis for Who/When, and generates role-aligned handoff notes for RD, UI/UX, and QA. Use when the user asks to clarify a requirement, write a PRD, do 5W2H analysis, define acceptance criteria, or align RD/design/QA on scope.
Fetch GitHub repository contents using uithub CLI when users paste GitHub URLs. Use when users share github.com links or need to analyze repository code, issues, or pull requests.
Set up automation — prospective memory triggers, neuro-symbolic rules, and CLAUDE.md sync. Use when the user says 'remind me when', 'trigger when', 'create a rule', 'auto-remember', 'sync to CLAUDE.md', 'push insights', 'set up trigger', 'when I open this file', 'when this keyword appears', or when you want to automate memory behavior based on conditions.
Track multi-session work with beads_rust (br) issue tracker, including creating/updating/closing issues, managing dependencies, syncing JSONL/DB, configuring prefixes, and migrating bd workflows to br. Use when a project uses .beads/ with br or when converting bd instructions to br.
Use when AWS CLI commands fail with SSO token expiration errors like "Token has expired", "SSO session has expired", or "Error when retrieving credentials".
This skill should be used when the user asks to "create agent", "build agent", "new agent", "add agent", "에이전트 만들어줘", "에이전트 추가해줘", "서브에이전트 추가", or when the orchestrator detects no suitable agent exists for a task.
Launch Claude Code via mclaude in headless (non-interactive) mode, bypassing the TUI. Use this skill when the user asks to run Claude Code with a specific provider, when automating Claude Code invocations, or when scripting with mclaude. Also use when the user asks how to use mclaude from the command line without the TUI, or wants to know available providers and models.
Search Confluence pages and fetch page content. Use when a user wants to find or read Confluence wiki pages, or when context from Confluence is needed.
Transforms AI-generated text into natural human writing by detecting and removing 37 AI patterns, injecting authentic voice, and varying rhythm. Use when text sounds like a chatbot wrote it, when preparing content for publication, or when AI detection scores need to drop.
