oh-my-superpowers

Generate a daily AI tech brief: scan configured sources (X accounts, blogs, tech communities), filter high-value items, and render a card-based HTML newsletter. Use when the user asks for a daily AI digest, tech briefing, information scan, or says "今日简报", "AI 日报", "tech brief", "扫一遍今天的信息". Do NOT use for deep research on a single topic (use deep-research), or for X account content curation without the brief format.

Generate a daily GitHub trending project-intelligence report: fetch trending repos, study README/API evidence, answer each repo's purpose/content/value in Chinese, and publish a lens-first HTML digest to html-serve. Use when the user asks for GitHub trending, hot repos, 今日 GitHub 热门, trending 项目报告, or a daily open-source project digest. Do NOT use for deep research on a single repo (use deep-research) or general AI news briefs (use daily-ai-brief).

Use when designing a polished static HTML page prototype and its DESIGN.md: clarify the user's page scenario, choose an information organization model, search DESIGN.md references, create multiple adjustable prototypes, publish them to html-serve, and iterate from user feedback. Do NOT use for routine publishing of an already-approved HTML template.

Use when turning one or more IMAP mailboxes into structured local outputs: classified JSONL events, downloaded attachments, and auditable processing state. Use for repeatable mailbox ingestion, classification, and attachment archiving SOPs. Do NOT use for sending mail, MCP email tools, one-off email reading, or full email-client workflows.

Use Web Operator to inspect, navigate, extract, and run site-specific SOP workflows in a local Chrome-family browser. Use when the user needs a workflow on a live Chrome tab, a bundled site workflow script should run, or a page already open in Chrome should be inspected or debugged.

Create a new Agent Skill from a requirement — grounding on what a skill is, clarifying what to build, picking a skill pattern, drafting SKILL.md plus references, and reviewing the result. Use when the user says "写个 skill", "创建 skill", "把这个能力做成 skill", "build a skill", "make a skill", or asks to author or scaffold a skill, even without naming the format. Do NOT use for restructuring an existing skill (skill-refine), auditing a finished skill (skill-review), or non-skill Markdown.

Use when a task requires systematic multi-round research across multiple angles, sources, and validation steps before producing a conclusion or report. Do NOT use when a quick ad-hoc search, a single-page summary, or a one-off fact lookup is enough.

Review, audit, and improve an Agent Skill for spec compliance, design quality, expression quality, and eval readiness. Use when reviewing a skill directory, checking a SKILL.md file, auditing references or scripts, diagnosing why a skill triggers poorly, or evaluating whether a skill is ready to deploy — even if the user doesn't explicitly say "skill review." Do NOT use for reviewing agents (use agent-review instead).

Use when an agent needs lightweight structural navigation of a local codebase: index JS/TS/Python/Bash projects, find symbols, inspect callers/callees, retrieve snippets, and check whether an index is stale.

Use when managing a unified local invoice registry with `omp invoice`: scheduled scans of configured local source directories, pending invoice extraction, ownership assignment, claim/substitute classification, status updates, listing, and archive operations. Do NOT use for downloading email, reading mailboxes, or source-specific chat app automation.

Interview the user relentlessly about a requirement, plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to clarify a requirement, stress-test a plan, get grilled on their design, or mentions "grill me".

Build a documentation skeleton with maintenance contracts for a project that already has running MVP code but lacks a sustainable doc structure. Use when scaffolding project docs (PROJECT/LANGUAGE/PRODUCT/DESIGN/ architecture/ADR), adding contract frontmatter to existing docs, or migrating from MVP doc state to long-term form. Do NOT use for brand-new empty projects, design dialogue, code architecture refactoring, or global doc/code drift audits.

Use when the user is ready to ship a clarified requirement and needs state that survives `/compact`. Trigger only when the user explicitly types `/kickoff` or says "kickoff". Skip when the requirement is still ambiguous — kickoff does not clarify requirements.

Systematic debugging for failing tests, runtime errors, and unexpected behavior. Use when investigating bugs, finding root causes, or before attempting fixes. Do NOT use for feature design, task orchestration, or speculative rewrites.

Design-first collaborative dialogue that turns a fuzzy requirement into an approved spec before implementation begins. Trigger only when the user explicitly types `/brainstorm` or says "brainstorm", "brainstorming", or "头脑风暴". Skip for trivial single-file changes, factual lookups, or open chats with no design intent — direct coding handles those.

Review local code changes: uncommitted work (staged/unstaged) or recent unpushed commits. Use when code needs quality review during local development. Do NOT use for PR review or branch comparison.

Multi-perspective roundtable debate using multiple AI runtimes (claude/codex/pi) with historical figure personas. Use when a topic benefits from adversarial multi-viewpoint discussion.

Multi-runtime orchestration patterns and prompt templates for composing pipelines / fan-out / discussion / batch workflows over `omp dispatch`. Use when planning multi-step or multi-agent workflows across claude/codex/pi. Do NOT use for single one-off calls (call `omp dispatch run` directly).

Use before running /compact to preserve session context. Generates a structured .handover.md and a ready-to-use /compact instruction so the next session resumes without information loss. Trigger when context reaches ~65-70%. Do NOT use for general summaries or task documentation.

Use when building or maintaining an LLM-powered knowledge base with `omp wiki`. Triggers: 'add to wiki', 'ingest this article/URL', 'what do I know about', 'search my wiki', 'lint my wiki', 'compile wiki', 'knowledge base', '加入知识库', '存入 wiki', '查一下我的 wiki', or any mention of 'llm wiki' or 'omp wiki'.

Use when writing or editing any Markdown file. Helps choose document structure, heading hierarchy, wording strength, and the right Markdown medium such as lists, tables, diagrams, and tag blocks. Includes focused guidance for skill documents and architecture documents. Do NOT use for non-Markdown files, code quality work, or factual research that should happen before writing.

Rewrite Markdown files inside a skill directory to improve structure, hierarchy, clarity, concision, expressive strength, and directory-level consistency. Use when a skill's SKILL.md, references, scenarios, or Markdown assets feel structurally messy, verbose, weakly expressed, or inconsistent. Do NOT use for script/code fixes, spec-compliance review, or non-skill Markdown projects.

Use when a task matches a pre-defined Pi Agent's expertise. Delegate via omp run instead of doing it yourself. Do NOT use for multi-agent orchestration (use team skill).

Review and audit a Pi agent markdown file for spec compliance and design quality. Use when: reviewing an agent file, checking if an agent is ready to deploy, auditing agent description or system prompt quality. Do NOT use when: reviewing a skill directory (use skill-review), designing a new agent (use brainstorming), or testing agent runtime behavior.

Evolve project skills and CLAUDE.md based on real usage data. Use when you want to audit and improve skills in the current project using cross-project session analysis, user feedback from memories, and rule consistency checks. Do NOT use for one-off skill fixes or new skill creation.

项目记忆系统。回忆/回顾经验教训(recall)，提取/记住对话经验(capture)， 提炼洞察模式(evaluate)，查看记录(list)。

Use when converting local Markdown files into publishable outputs. (1) "转成PDF"/"导出PDF"/"转成图片"/"导出PNG"/"导出文件" (2) "高质量图片"/"卡片风格"/"海报"/"好看一点"/"设计感"

Use when the media-editor agent needs to initialize the data directory, save archived media items, query the archive, or promote items to root-archive. Do NOT use when browsing the web, generating reports, or doing tasks unrelated to media archive management.